The Vanguard Real Estate ETF (NYSEARCA:VNQ) is the default way most investors get real estate exposure, and for good reason. It holds 159 real estate positions across $38.2 billion in assets, charges an expense ratio of just 0.13%, and pays a 3.57% dividend yield. The fund has climbed 13.4% year to date, which sounds respectable until it is set against the slice of the real estate market that is actually pulling the sector higher. Investors owning VNQ for its yield and diversification may be underweighting the one theme driving the sector’s returns: data-center REITs powering artificial intelligence.
Why VNQ’s Broad Exposure Is Diluting the AI Trade
VNQ owns every corner of the listed real estate. That includes healthcare landlord Welltower at 7.65%, logistics operator Prologis at 7.14%, and retail names like Simon Property Group at 3.56%. Data-center and tower operators are represented, but sparingly: Equinix sits at 5.63%, American Tower at 4.65%, and Digital Realty at 3.49%. In practice, the AI infrastructure theme accounts for a low double-digit share of the fund.
The macro backdrop reinforces the point. Residential construction is cooling, with housing starts falling 15.4% month over month to 1.18 million in May 2026, and existing home sales stuck at 4.17 million annualized. Real estate value added in GDP grew just 1.1% in Q1 2026, while the information sector grew 1.5% and hit 3.2% in Q3 2025. VNQ ties investors to the slower half of that split.
The Alternative: Global X Data Center & Digital Infrastructure ETF
The Global X Data Center & Digital Infrastructure ETF (NASDAQ:DTCR) focuses on the real estate segment that is actually compounding. The fund holds $1.22 billion in net assets, anchored by Equinix at 13.80%, Digital Realty at 12.61%, American Tower at 12.25%, and Crown Castle at 8.58%. It supplements those REIT positions with pure-play AI operators, including GDS Holdings at 3.99% and Applied Digital at 3.21%, as well as semiconductor suppliers feeding the compute buildout.
The mechanism behind that outperformance is straightforward. Roughly 40% of DTCR’s book is the same handful of tower and data-center REITs held by VNQ, but at three to four times the weighting. That concentration flows directly into returns: DTCR is up 36% year-to-date and 56.52% over the past year, compared with VNQ’s 12.87% one-year gain. Anchor holding Equinix has delivered 33.96% year-to-date on its own. The underlying capex cycle shows why: hyperscaler spending on data-center capacity is projected to keep growing roughly 25% annually through the late 2020s.
The Tradeoffs Are Real
An expense ratio of 0.50% makes the fund meaningfully more expensive than the 0.13% charged by the Vanguard alternative, and its $2.57 billion asset base remains a fraction of the larger fund’s $38.2 billion. Concentration cuts both ways: the strategy is down 7.37% over the past month, while the broad sector gained 2.32%. The portfolio also blends semiconductor equities like NVIDIA, Broadcom, and AMD, ensuring it behaves more like a tech ETF than a pure REIT.
Income is another point of contrast between the two funds. VNQ pays a 3.57% yield, though the 104.33% payout ratio raises questions about coverage. DTCR is structured for capital appreciation, and its yield is far lower.
How to Approach the Switch
A full swap meaningfully changes the risk profile. A partial reallocation, replacing part of a VNQ position with DTCR, keeps broad real estate exposure while tilting toward the AI infrastructure theme. In a taxable account, VNQ shares purchased before the 2025 rally likely carry embedded gains that would trigger taxes on sale. Tax-advantaged accounts avoid that friction. Investors already holding NVIDIA, Broadcom, or other AI names elsewhere should note that DTCR overlaps with those positions.
What This Adds Up To
The broad real estate fund remains a sensible core holding thanks to its extensive diversification and low fees. Its structure inevitably dilutes the focused exposure to the specific theme driving current returns, whereas the digital infrastructure alternative provides that concentration at the cost of elevated volatility, a smaller asset pool, and integrated semiconductor risk. Investors whose original thesis for owning the broad fund was participation in the AI-infrastructure buildout rather than traditional property ownership may find this mismatch worth resolving.
Contact [email protected] for any questions or corrections.