Equinix, Inc.

▲ Bull Case

• ◆Platform re-rating: P/AFFO expands from 25x to 28-30x as EQIX is framed as neutral infrastructure (Visa/CME comparison) rather than commodity REIT; AFFO $51-53/share FY2028 × 29x = $1,479-1,537 (+40-45%)
• ◆Interconnection pricing power inflection: EQIX prices cross-connects at value-based rates with zero substitutes among 5,800+ connected parties; interconnection revenue grows 15%+ vs. 9% base; margins expand 200-300bps; each $100M incremental interconnection revenue = ~$1/share AFFO upside
• ◆xScale JV $5B+ new commitment and cap rate compression: New sovereign capital partner adds $5B+; AI demand drives data center cap rate compression from 5.5% to 4.5%; EQIX's 25% equity stake appreciates; JV management fees increase; AFFO upside $2-4/share not yet in guidance

▼ Bear Case

• ◆10yr UST rises above 5% + hyperscaler bookings digestion: Rate shock causes broad REIT de-rating to 18-20x P/AFFO; simultaneously hyperscalers complete major builds and reduce colo footprint in FY2027; bookings fall to $250-300M/quarter; AFFO growth slows to 5-7%; AFFO $43 × 20x = $860 (-19%)
• ◆AI capex digestion air pocket in FY2027: Cloud giants absorb in FY2027 the AI infrastructure pre-committed in FY2025-2026; new bookings slow sharply for 4-6 quarters; management lowers FY2027 guidance mid-year; stock falls 20-25% before bookings cycle resumes
• ◆Margin compression from power cost inflation: Electricity and cooling costs rise faster than CPI in constrained markets (Northern Virginia, UK, Singapore); power purchase agreements expire and renew at higher rates; Adj EBITDA margin fails to expand as expected; AFFO growth undershoots guidance by 200-300bps/year

“Is EQIX's interconnection moat durable enough to justify a platform premium, or is it a REIT with cyclical AI bookings at a full multiple? Bull: The interconnection layer is structurally unreplicable—building a competing neutral network in 70 countries with 5,800+ connected parties takes 20+ years and tens of billions. Enterprises commit to Equinix Fabric for years or decades (high switching costs). AI inference requires edge proximity that only Equinix's density provides. At 25x P/AFFO, the market prices this as a REIT, not the platform it is. Bear: Growth depends on physical assets in specific locations; power constraints can impair growth trajectory. Hyperscalers build private interconnection (AWS Direct Connect, Azure ExpressRoute) that partially substitutes for EQIX cross-connects. The AI surge is a 2-3 year phenomenon, not a permanent re-rating driver. Our view: The bull is correct that the network effect is durable, but the bear is right that $1,059 is not cheap enough for high-conviction buy. HOLD and ACCUMULATE on pullbacks; the platform re-rating is a real 3-5 year possibility, not yet a 2026 certainty.”

• ◆Q1 2026 bookings ≥$400M/quarter with 60%+ AI revenue (confirms AI momentum is structural, not spike)
• ◆xScale JV new $5B+ commitment with expanded footprint (validates hyperscale thesis; platform acceleration)
• ◆Interconnection revenue acceleration to >10% YoY growth (demonstrates pricing power; supports platform re-rating narrative)
• ◆FY2026 AFFO guidance raise above midpoint (execution on record bookings; margin expansion confidence)
• ◆Rate cut: 10yr UST falls below 4.0% (REIT sector re-rating; multiple expansion for growth REITs)

• ◆AI bookings air pocket in FY2027 as hyperscalers digest FY2025-2026 capex (25% probability; moderate impact; mitigated by long-term contracts and recurring base)
• ◆Power/permitting constraints in Northern Virginia and UK limiting capacity growth (25% probability; moderate impact; creates pricing power on existing capacity)
• ◆Hyperscaler self-build reducing EQIX colo demand (20% probability; moderate impact; mitigated by interconnection moat and xScale JV structure)
• ◆Rate shock: 10yr UST sustainably above 5.5% causing broad REIT de-rating (15% probability; moderate impact; AFFO yield of 4% becomes uncompetitive vs. treasuries)
• ◆CEO Fox-Martin execution or strategic deviation from interconnection-first model (20% probability; low-moderate impact; CFO continuity and structural bookings momentum de-risk)