Data Centers & Cloud Infrastructure

Investment Thesis

Migration from on-premise compute to public cloud infrastructure is roughly 40-50% complete globally, leaving a long runway for continued workload transition. AI inference and training require always-on, low-latency infrastructure that favors large-scale providers over edge deployments. Colocation REITs serve enterprises with data sovereignty or hybrid requirements, benefiting from the same underlying demand drivers without direct AI model risk.

Theme Score Breakdown

Data Centers & Cloud Infrastructure sits at the bottom of Strong because cloud migration has a measurable multi-year runway and AI workloads require always-on infrastructure that only hyperscalers can deliver at scale. Revenue from AWS, Azure, and GCP is among the most precisely reported capex data of any theme. The score is held at the boundary of Strong because approximately half the universe overlaps with AI Infrastructure — the most exciting growth is captured in that theme — and the cloud migration story is gradually becoming a mature, competitive market. Policy support is limited (limited direct subsidies; active antitrust headwinds).

Key Risks

• Hyperscaler capex cycles causing periodic overbuild and utilization pressure
• Regulatory antitrust scrutiny on cloud market concentration
• Energy costs and power availability constraining data center margins
• Sovereign data laws fragmenting global cloud architectures