AI-driven infrastructure convergence reshapes real estate investment focus in 2025

As compute demand from artificial intelligence accelerates, the physical architecture that supports it is reordering investor priorities. What began in 2025 as a wave of hyperscaler announcements for multi‑gigawatt data‑centre buildouts has expanded into a broader thesis: the convergence of compute, connectivity and energy into a single, investable layer of the economy. According to WisdomTree, that convergence is the foundation of the WisdomTree New Economy Real Estate Fund (WTRE), which reframes “real estate” to include data centres, telecom towers, edge nodes, modern logistics and even satellite networks that together form the substrate of large‑scale intelligence.

A structural pivot, not a cosmetic tweak
WisdomTree describes WTRE’s rules‑based index as selecting companies that derive at least half their revenue, profit or assets from digital and industrial economy infrastructure. The Index allocates 90% to established digital and industrial infrastructure and 10% to next‑generation digital infrastructure, a split intended to balance scale and innovation. According to WisdomTree, that design has led the Fund to overweight legacy leaders such as Equinix, American Tower and Prologis while also including frontier operators such as Iris Energy, Applied Digital and AST SpaceMobile.

The pivot has been explicit and dated. WisdomTree notes that WTRE moved away from its previous ex‑U.S. real estate benchmark in April 2022, adopted a New Economy remit, and on 10 October 2025 began tracking the WisdomTree New Economy Real Estate Index’s total return performance. That structural change, the firm says, underpins the Fund’s reorientation from income‑driven, traditional REIT exposure to a growth‑oriented, technology‑anchored real assets strategy.

Performance and market reception
The shift has translated into markedly different outcomes. WisdomTree reports that WTRE’s one‑year return of 13.79% and three‑year annualised return of 14.76% each ranked the Fund fourth among more than 150 managers in the Morningstar Global Real Estate peer group as of end‑September 2025, placing it ahead of 98% of peers for those periods. WisdomTree also highlights the contrast with the Fund’s legacy performance: a 10‑year annualised return of 2.85% that reflected exposure to traditional, non‑U.S. real estate during an era of rising rates and weakening office fundamentals.

Independent market trackers show continued investor interest. StockAnalysis reports the London‑listed UCITS ETF version of WTRE returned 26.83% total for the 12 months to 19 December 2025, and its holdings data indicates a diversified roster of roughly 58 names led by logistics and infrastructure incumbents such as Prologis and Goodman alongside data‑centre and tower owners including Equinix, Digital Realty, Crown Castle and American Tower, plus newer entrants such as AST SpaceMobile. Analyst consensus compiled by TipRanks gives WTRE a “Moderate Buy” rating, with an average 12‑month price target that signals upside relative to current levels.

Beyond data centres: a systems view of infrastructure
WisdomTree and sector research urge investors to look beyond roof‑top compute. Industry data shows vacancy rates in North American data centres tightened in H1 2025 as hyperscaler and AI demand consumed available capacity, a dynamic cited by CBRE Research that supports the argument demand is outpacing supply. But the constraint is broader than rack space: every additional gigawatt of AI capacity requires transmission lines, substations, fibre, cooling and interconnection capacity. That creates opportunity across subsectors, energy procurement and storage, subsea and terrestrial fibre, tower densification for edge connectivity, and LEO satellite systems, that together reduce latency, increase resilience and bring compute closer to users and devices.

Companies that once occupied niche corners of crypto or telecom infrastructure are recasting themselves for the AI era. WisdomTree highlights Iris Energy’s evolution from Bitcoin mining to multi‑gigawatt, renewables‑aligned compute campuses; Applied Digital and Cipher Mining’s moves to repurpose blockchain‑era facilities into multipurpose compute campuses; and AST SpaceMobile’s commercial steps toward space‑based broadband as examples of firms bridging energy, compute and connectivity. According to WisdomTree, these stories illustrate a wider shift: winners in the decade ahead may be those who control the “pipes and power” as much as the silicon.

Risks, valuation and investment implications
WisdomTree acknowledges the exuberance around next‑gen infrastructure and the attendant hazards: elevated valuations, capital intensity and the necessity of securing long‑term, low‑cost power and fibre. The firm frames WTRE’s index methodology, semi‑annual reconstitution, free‑float adjusted market‑cap weights and a 90/10 stability‑innovation balance, as a guardrail against one‑off enthusiasm, but also as an explicit bet that the physical layers of the internet will capture a growing share of real‑asset value.

Investors should weigh three practical implications. First, the AI buildout creates diversified exposure opportunities beyond hyperscaler equities, spanning REITs, specialised infrastructure operators and satellite and tower companies. Second, the constraints to adding incremental compute, energy, land and interconnection, mean future returns will depend materially on execution and contracts that secure power and capacity. Third, while past underperformance of legacy REIT exposures explains some of WTRE’s repositioning rationale, historical returns remain an imperfect guide; the Fund’s recent rankings reflect exposure to an active sectoral rerating that may be cyclical as well as structural.

Conclusion
The narrative of digital infrastructure in 2025 has shifted from racks and chips to systems engineering: power grids, fibre and orbital assets as much as GPUs. According to WisdomTree, WTRE is built to capture that systems view by redefining real estate to include the physical backbone of AI and cloud economies. Independent data and market trackers show investors have rewarded that repositioning to date, but the landscape remains capital‑intensive and execution‑sensitive. For allocators, the choice is no longer whether to own technology exposure, but how to own the physical layers that will determine which technology firms can scale their intelligence most efficiently.