By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panelists generally agree that while AI-driven demand has boosted data center REITs, high valuations, capex intensity, power constraints, and interest rate sensitivity pose significant risks. The potential for hyperscalers to vertically integrate and build their own data centers is a key concern.
Risk: Hyperscalers accelerating in-house data center builds and potential power grid constraints in Northern Virginia
Opportunity: Realty Income's 'triple net' approach and potential resilience in a scenario where hyperscalers move in-house
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Real estate investment trusts (REITs) have staged a comeback in 2026 after years of declining or moving sideways. And one category of REITs, data centers, is particularly hot right now.
But first, what are these investments, and how do they work? REITs are companies that finance or own, and often also manage, income-producing real estate. The properties they own can vary widely, from office and retail properties and residential real estate to healthcare and data center assets, among many other categories.
Missed Nvidia in 2009? This Rare Signal Is Flashing Again. In 2009, a "Double Down" signal flashed for a little-known chipmaker called Nvidia. For the first time in years, that same "Total Conviction" signal is flashing for a company 1/100th the size of Nvidia. Continue »
Yet the entire REIT sector struggled in the aftermath of the COVID-19 pandemic due to trends such as work-from-home culture and depressed spending at brick-and-mortar retail outlets. Plus, the Federal Reserve's elevated interest rates, implemented after the pandemic to counter rising inflation, hurt REITs, many of which rely heavily on debt to expand.
But REITs have staged a comeback in 2026, as those post-COVID trends have waned and interest rates have stabilized. The entire REIT sector is up more than 10% year to date, as measured by the Vanguard Real Estate Fund ETF (NYSEMKT: VNQ). That's about the same as the broader market -- the S&P 500 is up 10.5% so far this year. In fact, REITs were the best-performing asset class in June.
Yet a few specific REIT categories are doing far better, with returns that have wildly outpaced this year's returns. Among them is a relative newcomer to the REIT space: data centers. The year-to-date return for data center REITs was 33.2% as of June 30, according to the National Association of Real Estate Investment Trusts. That's been driven by extraordinary growth in the sector, particularly from AI companies and rising data usage.
Two well-known data center REITs, Equinix (NASDAQ: EQIX) and Digital Realty Trust (NYSE: DLR), are up 34% and 14%, respectively, year to date.
And there's a new entrant to the space: Realty Income (NYSE: O). Realty Income recently declared its 673rd consecutive monthly dividend. The REIT has increased its dividend 135 times since it went public in 1994. Its dividend yield is a hefty 5.08%.
Realty Income's portfolio of properties spans the retail, industrial, and gaming sectors. But now it's getting deeper into data centers. The company recently announced a $6 billion joint venture with Cloud Capital that will invest in a diversified portfolio of stabilized hyperscale assets leased to investment-grade tenants under long-duration, triple net leases (a commercial lease agreement in which the tenant agrees to pay all property operating expenses, like taxes and insurance, in addition to rent).
The new venture's first investments will be three facilities in Northern Virginia's "data center alley."
If you believe the data center build-out will continue and want to profit from it, consider a $1,000 investment in one of these three stocks today.
Before you buy stock in Realty Income, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Realty Income wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $400,964! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,272,955!
Now, it’s worth noting Stock Advisor’s total average return is 930% — a market-crushing outperformance compared to 210% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
**Stock Advisor returns as of July 17, 2026. *
Matthew Benjamin has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Digital Realty Trust, Equinix, Realty Income, and Vanguard Real Estate ETF. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Four leading AI models discuss this article
"Data-center REIT momentum is real but already priced in at current multiples; selective entry only after confirming hyperscaler lease momentum and power availability beyond 2026."
The article correctly flags the AI-driven tailwind for data centers, with EQIX +34% and sector REITs +33% YTD versus VNQ’s 10%. Realty Income’s (O) $6B JV into hyperscale triple-net assets in “data center alley” diversifies its retail-heavy portfolio and should support its 5.08% yield. However, the piece glosses over capex intensity, power constraints in Northern Virginia, and rising interest-rate sensitivity for REITs still carrying elevated debt loads post-2022. Valuation also matters: EQIX and DLR trade at 22-25x forward FFO while O sits at ~14x; any slowdown in AI capex or hyperscaler lease renewals could compress multiples quickly.
AI build-out hype could stall if GPU utilization disappoints or enterprises delay private-cloud migration; a renewed spike in Treasury yields would hammer REIT debt costs and send O, EQIX, and DLR sharply lower regardless of occupancy.
"Realty Income's entry into data centers is a yield-preservation strategy that shifts their risk profile toward infrastructure-heavy capital expenditure rather than pure retail stability."
The pivot to data centers by Realty Income (O) is a defensive move to combat the stagnation of their traditional retail portfolio. While the 5.08% yield is attractive, investors must distinguish between pure-play data center REITs like Equinix (EQIX) and Realty Income's 'triple net' approach. Realty Income is essentially acting as a landlord for infrastructure, not a tech-growth play. The 33.2% YTD return in the sector is impressive but reflects a massive valuation expansion that may be pricing in perfection. If AI demand faces a cooling-off period or power grid constraints in Northern Virginia stall expansion, these REITs will face severe multiple compression given their sensitivity to interest rate volatility.
If the AI infrastructure build-out becomes a multi-decade utility-like necessity, the 'triple net' lease structure provides a recession-proof cash flow that makes these REITs the ultimate bond proxy in a high-growth sector.
"Data center *demand* is real; data center REIT *valuations* already price in a decade of flawless execution at premium multiples, leaving little margin for error."
Data center REITs are up 33% YTD on genuine AI capex tailwinds, but the article conflates two separate theses: (1) data center *fundamentals* are strong—true, with hyperscalers locked into long-duration leases at rising rents; (2) data center REIT *valuations* are attractive—highly questionable. EQIX trades ~28x forward P/E; DLR ~22x. Both assume sustained 15%+ annual rent growth. The article never mentions cap rates, debt levels, or refinancing risk. Realty Income's $6B JV is a footnote, not a core business pivot—retail/industrial still dominate its portfolio. The Netflix/Nvidia hindsight marketing is noise.
If AI capex peaks in 2027–2028 and hyperscalers build in-house, utilization rates collapse and rents compress sharply. REITs with floating-rate debt face margin squeeze in a higher-for-longer rate environment.
"Even with AI-driven demand, a looming capex cycle and rate sensitivity may cap upside in data-center REITs, making Realty Income's hyperscale pivot a higher-risk diversification than the article implies."
AI-driven demand has buoyed data-center valuations, but the article glosses over key risks. Realty Income is a traditional retail REIT pivoting into hyperscale data centers via a $6B JV with Cloud Capital, which introduces capital-intensity, rate sensitivity, and operator risk into a portfolio that has been built on long-term triple-net leases. The Northern Virginia data center alley is crowded; supply surges could pressure rent growth and cap rates if cloud capex slows. Higher financing costs, energy price volatility, and regulatory/green-energy pressures could squeeze cash flow. For all the hype, a meaningful re-rating requires sustained AI demand and tight capacity—uncertain in the near term.
Counterpoint: a data-center glut and higher financing costs could cap rent growth and delay ROI. If hyperscale underwriting proves tougher than expected, Realty Income’s pivot could underperform.
"Hyperscaler vertical integration poses a larger existential risk to pure-play data-center REITs than near-term power or rate constraints."
Nobody has flagged the second-order effect of hyperscalers vertically integrating. If Microsoft, Google, and Amazon accelerate in-house builds beyond 2026, EQIX and DLR’s lease-up assumptions collapse faster than cap rates can adjust. Realty Income’s triple-net JV may actually prove more resilient here, acting as a pure pass-through with lower execution risk than the pure-plays.
"Regulatory and power-grid constraints create a durable moat for incumbent data center REITs that vertical integration by hyperscalers cannot easily bypass."
Grok, your pivot to vertical integration is the real elephant in the room. However, you are underestimating the regulatory and power-grid moat. Hyperscalers can build the hardware, but they cannot easily replicate the land-use permits and 500MW+ utility interconnections already secured by DLR and EQIX. Realty Income’s JV is a capital-allocation play, not a moat-building one. If the grid becomes the primary bottleneck, the incumbent REITs hold the keys, not the tenants.
"Grid bottlenecks favor energy producers and hyperscalers with capital, not landlords."
Gemini's grid-moat argument overstates EQIX/DLR's defensibility. Hyperscalers are already securing their own interconnections—Microsoft's Puget Sound deals, Google's Iowa buildouts. The real constraint is *power generation*, not permits. REITs own real estate; they don't own electrons. If capacity tightens, hyperscalers build captive solar/nuclear or relocate. Realty Income's JV actually hedges this better: it collects rent regardless of who solves the energy puzzle.
"Even with potential hyperscaler in-house builds, financing costs and energy prices dominate risk to data center REIT multiples, not just occupancy."
Grok's concern about hyperscalers moving in-house is worth flagging, but the real swing factor is financing and energy costs, not just tenants. Even if self-build rises, hyperscalers still need dense interconnection hubs and carrier-grade campuses—the moat shifts toward ecosystem value, risk management, and capex discipline. If rates stay high and capex inflation persists, EQIX/DLR could see multiple compression even with steady occupancy.
The panelists generally agree that while AI-driven demand has boosted data center REITs, high valuations, capex intensity, power constraints, and interest rate sensitivity pose significant risks. The potential for hyperscalers to vertically integrate and build their own data centers is a key concern.
Realty Income's 'triple net' approach and potential resilience in a scenario where hyperscalers move in-house
Hyperscalers accelerating in-house data center builds and potential power grid constraints in Northern Virginia