DLR · Q3 2025 Earnings

Narrative arc: Mix shift to 0-1MW + interconnection → Funding model resolved via hyperscale fund → Backlog visibility into 2026-2027 → Connectivity as the AI moat.

The most consequential multi-quarter shift is how management frames the competitive landscape. Last quarter the message was that capacity-block discussions had moved to "late 2026 and early 2027 deliveries" — visibility but still implicitly a race for power. This quarter the framing escalates to a defensible moat: "It is harder to build in these locations for a growing list of reasons, and we expect this capacity will continue to be highly sought after." That shift — from "we have the power" to "nobody else can build where we are" — recasts DLR as a scarcity asset rather than a capacity competitor with the gigawatt-campus announcements.

The AI narrative has also tightened. Two quarters of "AI is a major driver" has become a more specific structural claim: AI training is moving to remote campuses, but inference workloads — which will dominate as models hit production — require low-latency metro proximity. Management's anchor: "even without a record hyperscale lease, like the one we signed in March of 2025, 50% of our bookings were related to AI use cases in the third quarter." This matters because last quarter the enterprise AI contribution was described as "modest" and "patient"; this quarter it is the supporting evidence for the metro-inference thesis. The customer base is doing the work of validating the thesis, not management's prospective framing.

Third, on funding capacity. Q2's tone resolved the balance-sheet objection via the U.S. Hyperscale Fund. Q3 takes the next step — leverage actually fell to 4.9x from 5.1x, and the Rollins exchange has management explicitly comfortable scaling on-balance-sheet builds before drawing the fund. Combined with $1.5B of stabilized assets still to contribute to the fund and a $1.3B January debt maturity at 2.5% rolling, the capital stack is being reframed as a strategic choice rather than a constraint.

Fourth, the guide-raise cadence itself signals a change in posture. After Q2's deliberate "growth will be best measured in years" expectation-flattening, Q3 abandons that restraint: the FY core FFO range was raised by more at the high end (+$0.13) than the low end (+$0.07), widening the range upward rather than narrowing it. Coupled with the new "expected to begin 2026 with significant momentum" line, management is now signaling rather than tempering.

Finally, on renewal spreads, management lifted both the cash and GAAP bands rather than holding or withdrawing them — an explicit endorsement of the supply-constrained pricing environment Mercier called out, and a tone-consistent finish to a fully bullish print.