EQIX Equinix

Research status: Primary-source verified through Equinix Q2 2026 results and current investor materials. Valuation basis: $1,065.39 close on 21 August 2026.

Equinix is a high-quality digital-infrastructure compounder rather than a pure AI beta. Q2 2026 revenue was $2.625B, +16% YoY, monthly recurring revenue grew 11% YoY, annualized gross bookings rose 23%, net interconnections increased by a record 9,700, and adjusted EBITDA margin reached 53%. AFFO/share was $11.78, +19% reported and +18% normalized/constant currency. Management raised 2026 revenue growth guidance to 11–12%, AFFO/share growth to 10–12%, and its 2027–2029 outlook to 10–13% annual revenue growth and 9–12% AFFO/share growth.

The AI angle is attractive because interconnection density, enterprise proximity and scarce power can benefit from distributed inference and hybrid architectures. But the capital requirement is material: 2026 total capex guidance is $5–6B, and the 2027–2029 framework assumes $5–7B annual capex. Equinix therefore needs to keep earning attractive returns on invested capital while protecting its balance sheet.

At $1,065.39, the stock deserves a quality premium but is not obviously cheap. PowerFund views EQIX as a lower-beta AI-infrastructure investigate whose role would be portfolio durability rather than maximum convexity.

Investment case

Equinix owns one of the world's densest networks of carrier-neutral data centers and interconnection ecosystems. Customers place workloads there because networks, clouds, enterprises and counterparties are already present, creating a location/network-effect moat that is difficult to reproduce.

> The PowerFund thesis is that AI, hybrid cloud and distributed inference increase the value of interconnection-rich digital infrastructure, allowing EQIX to compound recurring revenue and AFFO/share at high-single/low-double-digit rates while deploying large amounts of capital at attractive returns. The thesis fails if capex intensity rises faster than AFFO/share or if hyperscaler/self-build economics weaken Equinix's interconnection advantage.

Verified operating baseline — Q2 2026

• Revenue: $2.625B, +16% YoY.
• MRR growth: 11% YoY reported and normalized/constant currency.
• Annualized gross bookings: $424M, +23% YoY.
• Net interconnections added: 9,700, a record.
• Adjusted EBITDA: $1.396B, 53% margin.
• AFFO: $1.168B.
• AFFO/share: $11.78, +19% reported / +18% normalized constant currency.

2026 guidance

• Revenue: $10.205–10.285B.
• Adjusted EBITDA: $5.210–5.270B, ~51% margin.
• AFFO: $4.240–4.300B.
• AFFO/share: $42.69–43.29.
• Total capex: $5–6B.

2027–2029 management framework

• Annual revenue growth: 10–13%.
• 2029 adjusted EBITDA margin: 53%+.
• Annual AFFO/share growth: 9–12%.
• Annual capex: $5–7B.

KPI ladder

PowerFund valuation scenarios

These are PowerFund estimates, not company guidance.

24 months

• Bear (20%) — $800–900: growth slows toward mid-single digits, rates remain high and development returns compress.
• Base (55%) — $1,150–1,300: 9–11% AFFO/share compounding continues and valuation remains at a quality REIT premium.
• Bull (25%) — $1,450–1,600: AI/interconnection demand supports >10% AFFO/share growth, lower rates reduce financing pressure and xScale/enterprise bookings stay strong.

Probability-weighted midpoint is roughly $1,150–1,275/share.

60 months

• Bear: $850–1,000.
• Base: $1,550–1,900 with durable ~10% AFFO/share CAGR plus dividends.
• Bull: $2,200–2,600+ if AI/edge demand sustains elevated pricing and capital returns.

• Continued double-digit MRR growth and record/high bookings.
• AI inference and enterprise hybrid-cloud deployments increasing interconnection demand.
• xScale/JV expansion without disproportionate balance-sheet risk.
• AFFO/share growth staying at or above the new 9–12% long-term framework.
• Stable/declining financing costs.
• Scarce-power markets supporting pricing.
• Continued record interconnection additions.
• Margin expansion toward 53%+ by 2029.

• Capex intensity: $5–7B annual development spend must earn returns above cost of capital.
• Rate sensitivity: REIT valuation and financing costs remain sensitive to long rates/credit spreads.
• Hyperscaler self-build: large customers can build wholesale campuses directly.
• Power constraints: may delay projects or raise build costs.
• One-time items: Q2 included xScale fees that flattered reported growth; normalized metrics matter.
• Development risk: construction cost inflation and preleasing assumptions can reduce returns.
• Leverage/JV complexity: xScale structures, guarantees and off-balance-sheet exposures require monitoring.
• Competition: Digital Realty and regional providers compete for hyperscaler/enterprise workloads.
• Valuation: quality premium can compress even if fundamentals remain healthy.

Warning — freeze additions

• Normalized MRR growth <7% for two quarters.
• AFFO/share growth <5%.
• Adjusted EBITDA margin falls below ~50%.
• Interconnection additions turn negative or churn rises materially.
• Development yields approach cost of capital.
• Net leverage rises while bookings/occupancy weaken.

Reduce

• AFFO/share declines for two consecutive quarters.
• Annual capex remains >$6–7B while incremental AFFO growth slows materially.
• Major hyperscaler/self-build shift reduces xScale or interconnection economics.
• Persistent occupancy/churn deterioration across multiple metros.

Invalidate

> Invalidate if Equinix loses its interconnection-driven pricing/retention moat and can no longer convert heavy development spending into durable AFFO/share growth above its cost of capital.

Hard invalidators include sustained MRR growth <5–6%, structurally rising churn, or leverage/development returns deteriorating simultaneously.

Equinix competes on ecosystem density and interconnection, not lowest-cost bulk capacity.

• Digital Realty: closest global public peer with larger wholesale exposure.
• Hyperscalers: can self-build large campuses but still need neutral interconnection/enterprise proximity.
• Regional colo providers: can compete on price/local capacity but lack global ecosystem density.

The moat is strongest where customers value low-latency connectivity to many networks/clouds/enterprises in one campus. AI may strengthen this by increasing east-west data movement and hybrid inference needs, but wholesale AI campuses alone are not necessarily an EQIX moat.

1. Strip one-time xScale fees from quarterly growth/EBITDA comparisons.
2. Track same-store occupancy, churn, pricing and power pass-through by region.
3. Track xScale/JV commitments, guarantees and economics separately from owned IBX.
4. Estimate incremental AFFO per dollar of development capex.
5. Compare long-term return profile versus DLR and power-constrained infrastructure alternatives.
6. Monitor AI inference/enterprise workloads as distinct from hyperscale training campuses.
7. Identify valuation bands consistent with >12–15% expected annualized return including dividends.

Next full review: after Q3 2026 results or a material interest-rate / development-return change.