Digital Realty (DLR)

Revenue growing ~12-14% with same-store NOI growth accelerating to 8%+. Leasing volumes in 2025 were 2-3x historical averages. DLR is growing faster than its 5-year average and roughly in line with EQIX, though behind private hyperscale developers who can move faster on new capacity. The growth acceleration is structural, not cyclical.

DLR holds ~15% of the global third-party data center market. The PlatformDIGITAL strategy is attracting enterprise customers alongside traditional hyperscale tenants. However, hyperscalers (AWS, Google, Microsoft) are increasingly building their own facilities, and private developers (QTS/Blackstone, Vantage) are aggressively expanding. DLR's share is stable but the pie is growing faster than any single player.

Record leasing spreads of 15-25% on renewals reflect genuine supply-demand imbalance. Power constraints in key markets have shifted pricing power firmly to landlords. However, this pricing power is cyclical — once new supply comes online in 2027-2028, spreads will normalize. DLR is maximizing pricing in a favorable window.

DLR is expanding from traditional colocation/wholesale into AI-ready, high-density facilities (50+ kW per rack). The joint ventures with Blackstone and others accelerate development capacity. PlatformDIGITAL interconnection fabric adds differentiation. But the core product is still a data center — innovation is incremental.

Moving data center workloads is expensive, risky, and time-consuming. Customers build physical interconnections, establish network peering, and design around specific facility architectures. Churn rates are consistently below 5%. For enterprise colocation customers, switching costs are very high. For hyperscale wholesale, switching costs are lower since workloads are designed for portability.

DLR's interconnection-dense campuses benefit from genuine network effects — each additional customer increases the value for existing tenants through direct peering, reduced latency, and ecosystem density. The Ashburn campus is a global interconnection hub. These effects are location-specific and powerful but limited to major campus facilities.

Zoning and permitting for data centers is becoming more restrictive in key markets (Northern Virginia moratoriums, European environmental reviews). DLR's existing permits and entitled land are valuable. However, data center construction doesn't require proprietary technology — the barriers are capital and permitting, not IP.

Data centers require massive upfront capital ($1B+ per campus), creating barriers to entry for most players. But DLR competes against hyperscalers with unlimited balance sheets and PE firms with cheap capital. DLR's scale provides procurement advantages in power and equipment, but it's not a decisive edge in a capital-rich environment.

FFO estimates have been revised up ~10% over the past 6 months as leasing volumes and pricing exceed expectations. The street is modeling 10-12% FFO growth for the next 2-3 years. Revisions are consistently positive and accelerating — a strong signal.

Data centers are the consensus beneficiary of AI infrastructure spending. Every hyperscaler capex announcement is treated as bullish for DLR. The narrative is overwhelmingly positive, which is both good (momentum) and risky (crowded trade). Power constraint discussions add credibility to the supply-scarcity thesis.

CEO Andy Power has managed the portfolio well, improving tenant quality and accelerating development. The Blackstone JV was a creative way to fund growth without excessive dilution. Balance sheet is investment-grade. The concern is execution risk on a development pipeline that has doubled in 18 months — DLR has historically been a better operator than developer.