Two Landlords, Two Different Moats
They look like twins and aren't. The distinction drives everything:
- EQIX — Equinix. Colocation and interconnection. Equinix's edge isn't square footage — it's the dense ecosystem of networks, clouds and enterprises plugged into each other inside its facilities. Every new tenant makes the platform more valuable to the next one. That's a network-effect moat, and it produces sticky, high-margin, recurring revenue.
- DLR — Digital Realty. More hyperscale and wholesale. DLR leases big blocks of capacity to the cloud and AI giants. Bigger, more capital-intensive deals; more torque to the AI-capacity buildout; but also more concentration in a handful of enormous tenants and a heavier development bill.
Put simply: EQIX rents dense, connected space to thousands of stickier customers; DLR rents large-scale capacity to a few giant ones. One is a moat business; the other is a scale business.
Where the Prices Are
As of the August 24 close — note how mild these drawdowns are next to everything else in this series:
| Ticker | Price | Off 52wk high | vs 200-day | RSI(14) |
|---|---|---|---|---|
| EQIX | $1,055.20 | −6.5% | +11.0% | 49.0 |
| DLR | $188.55 | −9.4% | +6.0% | 46.7 |
RSI in the high-40s, single-digit drawdowns, both above trend. This isn't a momentum unwind — it's a mild pullback. The reason is structural: REITs were never priced like the 40x-growth contractors, so they had nothing like the same air to give back. That resilience is exactly what you're buying, and exactly what caps the upside. Hold that thought for the verdict.
What the Quarter Actually Said
Both delivered clean beat-and-raise quarters. REITs are measured on funds from operations (FFO/AFFO) per share, not EPS — it's the cash-earnings metric that adds back property depreciation.
| EQIX | DLR | |
|---|---|---|
| Revenue (YoY) | $2.63B (+16%) | $1.92B (+29%) |
| FFO/share (YoY) | $11.78 AFFO (+19%) | $2.13 core (+14%) |
| Bookings signal | $424M gross (+23%); +9,700 interconnects | $1.9B signed-not-commenced (record) |
| Guidance | Raised (largest ever) | Raised |
| Valuation | ~25–28x AFFO | ~22x FFO |
| Dividend yield | ~1.9% | ~2.7% |
| Net debt / EBITDA | ~5.3x | 4.7x |
DLR grew revenue faster (+29%), posted a record backlog of signed-not-yet-commenced leases, and — the standout detail — renewed expiring leases at +25% cash / +32% GAAP spreads. That pricing power tells you demand for existing capacity is white-hot. It trades cheaper (~22x FFO) and pays more (~2.7%), with lower leverage (4.7x). EQIX grew a bit slower on the top line but converted it into faster per-share cash growth (+19% AFFO), drove record interconnection bookings, and delivered what management called its largest guidance raise ever. It's the higher-quality, stickier revenue — and it carries a richer multiple (~25–28x) and more leverage (~5.3x) to match.
The Verdict — Which REIT
My pick between the two: Equinix (EQIX). When you're buying a landlord, you're buying durability, and EQIX has the better version of it. The interconnection moat makes its revenue stickier and higher-margin than wholesale space that has to be re-leased and re-priced; its per-share cash earnings are compounding faster; and it's less exposed to the concentration risk of a few giant hyperscale tenants. It's the higher multiple and higher leverage for a reason. If I only get one, I want the moat.
The higher-torque alternative: Digital Realty (DLR). If your specific goal is maximum exposure to the AI-capacity buildout, DLR gives you more of it, cheaper, with a bigger dividend and a cleaner balance sheet. Those +25–32% renewal spreads are a genuinely bullish signal on hyperscale demand, and the cheaper multiple gives you a little cushion. The trade-off is tenant concentration and a heavier capex bill. For an income-plus-AI-upside tilt, DLR is very reasonable — arguably the better value, if not the better business.
The Bigger Question: Landlord or Builder?
This is the one that matters for the whole series. Having now looked at every layer — the dirt (STRL), the systems (FIX, EME), the grid (PWR), the broader field (MTZ, MYRG and the rest), the hardware (ETN, VRT), and now the landlords — where does an investor actually want to sit?
The honest answer: the REITs are the conservative, lower-torque way to own AI, and the contractors are the aggressive, higher-torque way. The REITs collect durable, contracted rent with pricing power; that's why they barely fell. But they are also rate-sensitive (they borrow heavily, so higher-for-longer weighs directly on them), extremely capital-intensive (EQIX alone is guiding $5–7B of annual capex), and — this is the key point — they lag the builders in a capex supercycle. The contractor gets paid to build the center now; the landlord earns the rent slowly over the fifteen years after. In a boom, the builder's earnings inflect first and hardest.
So the framing I'd use: if you want to capture the buildout and can stomach 30–50% drawdowns like the ones this series just documented, the contractors (an EMCOR, a MasTec, an Eaton) give you far more torque per dollar. If you want to own the theme and sleep — collect a growing dividend, ride contracted rent, and take a fraction of the volatility — the REITs, and EQIX in particular, are the better vehicle. They're not competitors so much as two different risk settings on the same trade. The pullback we're living through is precisely the moment that distinction pays: the builders are on sale because they're volatile; the REITs held up because they're not.
The Headwinds That Could Break This
- Interest rates. The REIT-specific killer. These are leveraged, long-duration income vehicles; higher-for-longer rates raise their cost of capital and compress their multiples directly, regardless of how strong leasing is.
- Power and land constraints. The binding real-world limit on data-center growth is increasingly megawatts and interconnection, not demand. Both REITs can book leases faster than they can energize capacity — which paradoxically supports pricing (scarcity) but caps delivery.
- Capital intensity. $5–7B/yr of capex (EQIX) and DLR's heavy development pipeline mean growth is bought, not free. If the cost of that capital rises or AI leasing cools, the spend looks a lot less attractive.
- Hyperscaler concentration (DLR-specific). Leaning on a handful of giant tenants is great until one of them pauses. It's the same AI-capex-digestion risk as the rest of the complex, just expressed through a tenant roster instead of a backlog.
- The lag itself. If the AI buildout is shorter or shallower than hoped, the REITs — which monetize slowly — capture less of it than the builders did, having taken on the debt and capex up front.
Bottom Line
Between the two, I'd own Equinix (EQIX) for the interconnection moat and the higher-quality, stickier cash growth, and I'd treat Digital Realty (DLR) as the cheaper, higher-yielding, higher-torque hyperscale bet for anyone who wants more direct AI-capacity exposure with an income cushion. But zoom out: the REITs are the "own the theme and sleep" way to play AI, and this series' contractors are the "capture the buildout" way. The fact that the landlords barely moved while the builders got cut in half isn't a knock on the builders — it's the tell for what kind of investor you are. Decide which risk setting you want before you decide which ticker.
Prices and technicals pulled live from Yahoo Finance as of the August 24, 2026 close (market closed). Q2 2026 fundamentals from company releases and Nareit/Zacks/StockTitan coverage. FFO/AFFO multiples and leverage figures are approximate; EQIX's Q2 net debt/EBITDA is estimated from the prior quarter. Levels and multiples are point-in-time and will move.