AI companies need enormous amounts of physical server space, and Equinix is one of the biggest landlords of that space — it gets paid rent no matter which AI chip or software company wins. A professional investor publicly rebuilding a position suggests th
AI companies need enormous amounts of physical server space, and Equinix is one of the biggest landlords of that space — it gets paid rent no matter which AI chip or software company wins. A professional investor publicly rebuilding a position suggests they see the recent pullback as a buying opportunity rather than a warning. Data-center demand is contracted years in advance, which makes the revenue steadier than the headline AI chip names. Buying the landlord is a lower-drama way to play the AI boom as Nvidia earnings take the spotlight this week.
Idea
AI companies need enormous amounts of physical server space, and Equinix is one of the biggest landlords of that space — it gets paid rent no matter which AI chip or software company wins. A professional investor publicly rebuilding a position suggests they see the recent pullback as a buying opportunity rather than a warning. Data-center demand is contracted years in advance, which makes the revenue steadier than the headline AI chip names. Buying the landlord is a lower-drama way to play the AI boom as Nvidia earnings take the spotlight this week.
Advanced Analysis — institutional-depth research report
Verdict: the landlord thesis is sound, but the trend hasn't confirmed — wait
The 'buy the landlord' thesis is genuinely compelling on the fundamentals: EQIX grew revenue 42.1% to $9.22B in FY2025 with a 20.0% operating margin in the 91st percentile of its sector, and per the Yahoo Finance piece from August 24, 2026, SGA is rebuilding its Equinix position on accelerating AI demand. The strongest counterweight is that the landlord is not yet collecting — EQIX's free cash flow was -$400M (2nd percentile of its sector) with $4.31B of capex against $3.91B of operating cash flow, and leverage has climbed to roughly 1.38 debt-to-equity. Strategically, the setup is waiting for confirmation: ADX reads about 12.3-12.4 on both names versus the 20 threshold, and the 24-month backtest offers just one completed trade (+5.9% with a 7.96% max drawdown), so the sample is thin — and no robust parameter setup was established. Note also that a 0.71 correlation between EQIX and DLR means this basket is effectively one concentrated rate- and AI-capex-sensitive bet, not two. Verdict: wait for ADX above 20 or a tested bounce at first support (EQIX $1,051.92, DLR $190) before committing. **Conviction breakdown:** Thesis support 72 — contracted, durable revenue model with institutional validation, but leverage and cash burn temper it. Trade readiness 38 — entry conditions not live; ADX roughly 8 points below threshold. Risk quality 55 — defined 2.3% stop and structural invalidation levels are tight, but the two-name correlation (0.7089) concentrates risk. Backtest evidence 35 — a single completed trade over 24 months, and the 60-month window could not be evaluated due to a market-data gap. Fundamentals trend 78 — revenue and margin trajectories are strong; free cash flow is the outlier.
Trade now
The setup is not live today — it is waiting. Both legs closed Monday above their 21-day exponential averages (EQIX near $1,083 versus a $1,061 average; DLR near $194 versus $192), but the trend-strength filter is the binding condition: ADX (14) sits at 12.3 on EQIX and 12.4 on DLR, both needing a move above 20 to confirm a genuine trend rather than chop. The MACD lines are already above their signal lines, so that condition registers as near but a fresh crossover has not printed. On-balance volume versus price is not currently readable, so treat that condition as unverified until the feed resolves. The alternate entry is a support-bounce signal: the daily low touches or undercuts the nearest support (EQIX near $1,052, which also coincides with its 50-day average near $1,052; DLR at $190) and the close holds above it. EQIX is 2.8% above that support; DLR is 2.2% above its own. Risk is defined tightly. The hard stop is a 2.3% loss from entry, with take-profit at 4.6% — an effective reward-to-risk of 2:1 — and a structural stop if a daily close breaks below the second-tier support (EQIX near $1,044; DLR at $185). The 24-month backtest on this rule set produced one trade, a 5.9% return with a worst peak-to-trough drawdown of about 8%, which tells you this is a patient, low-frequency strategy. 'Wait' here means concretely: do nothing until either ADX clears 20 with the other crossover conditions intact, or price tags the nearest support and closes back above it. No robust parameter variant was established, so trade the rules as written
The AI landlord thesis is backed by real numbers
The idea's core claim — that Equinix is the steadier, rent-collecting way to play AI infrastructure — holds up well against the fundamentals. EQIX grew revenue 42.1% year-over-year to $9.22B in FY2025, placing it in the 65th percentile of its sector for growth. That is not chip-cycle revenue; it is contracted data-center capacity being monetized, which is exactly the durability the thesis argues for. Diluted EPS grew 58.3% to $13.76, showing the growth is reaching the bottom line. Profitability quality is equally supportive. EQIX's 20.0% operating margin ranks in the 91st percentile among information technology peers, and its 51.1% gross…
Scores
- Conviction score breakdown: 56
- Thesis support: 72
- Trade readiness: 38
- Risk quality: 55
- Backtest evidence: 35
- Fundamentals trend: 78
Watch items
- EQIX — ADX (14)
- DLR — ADX (14)
- EQIX — Nearest support (daily close above after a touch)
- DLR — Nearest support (daily close above after a touch)
- EQIX — Second-tier support (daily close below)
- DLR — Second-tier support (daily close below)
- EQIX — RSI (14)
- DLR — Price above EMA (21)
- DLR — MACD (12,26,9) crossed above MACD (12,26,9)
- DLR — ADX (14) above 20