Guidance
← Earnings Calls
01
02
03
04 05 06
Tone
AI
Capex
A1 A2 A3
DLR
FY2025 Q4
Improving
Digital Realty Trust, Inc. earnings call
Buzzberg read
Record backlog of $1.4 billion provides strong visibility.
Digital Realty delivered record FY2025 results with 10% core FFO growth and guided 8% growth for 2026. Demand remains robust across hyperscale and enterprise, with power availability as the primary constraint. The company launched a successful closed-end fund and is investing in AI inference-ready infrastructure. Record total bookings of $1.2B in 2025, with $400M in Q4; backlog reached $1.4B.
Buzzberg read Record backlog of $1.4 billion provides strong visibility. Digital Realty delivered record FY2025 results with 10% core FFO growth and guided 8% growth for 2026. Demand remains robust across hyperscale and enterprise, with power availability as the primary constraint. The company launched a successful closed-end fund and is investing in AI inference-ready infrastructure. Record total bookings of $1.2B in 2025, with $400M in Q4; backlog reached $1.4B. Read full analysisCollapse analysis
Digital Realty delivered record FY2025 results with 10% core FFO growth and guided 8% growth for 2026. Demand remains robust across hyperscale and enterprise, with power availability as the primary constraint. The company launched a successful closed-end fund and is investing in AI inference-ready infrastructure. Record total bookings of $1.2B in 2025, with $400M in Q4; backlog reached $1.4B.
- Zero-to-1MW interconnection leasing set records, driven by enterprise AI and network effects.
- Power availability is the key constraint; DLR leverages 5GW power bank and pre-emptive development.
- Private AI exchange platform gaining traction; inference demand expected to scale in 2026.
Revenue$1.7124B+9% QoQ
EPS$0.24Reported
Operating margin17.81%Reported
Free cash flow$0.719BReported
6 grounded callouts
What matters now
The highest-signal changes from the call.
Demand
Zero-to-one megawatt bookings set another record.
AI
Hyperscale leasing exceeded $800 million in 2025.
Show 3 more callouts
Guidance
2026 core FFO guidance implies 8% growth.
Capital Allocation
Private fund closed with $3.225 billion in commitments.
Capex
Development yields expected to remain in double digits.
Reported period
Actuals
| Metric | Reported | Change |
|---|---|---|
| Revenue | $1.7124B | +9% QoQ |
| EPS | $0.24 | Reported |
| Operating margin | 17.81% | Reported |
| Free cash flow | $0.719B | Reported |
| Capex | $0B | Reported |
| Net income | $0.0986B | +45% QoQ |
Forward-looking
Forward guidance
| Metric | Period | Range | Midpoint | Status |
|---|---|---|---|---|
| EPS | FY2026 | $7.90–$8.00 | $7.95 | Initiated |
AI, capex & demand read
Management read
Confident
Management emphasized record bookings, strong demand, and their ability to execute in a constrained supply environment, while maintaining a forward-looking positive outlook.
Management AI read
Management stated that AI adoption is accelerating and that inference demand will scale in 2026, positioning the company to benefit from increased enterprise AI workloads and interconnection demand. They also highlighted early adoption of their private AI exchange platform and their AI-ready infrastructure strategy with pre-installed liquid cooling and higher density deployments.
Investment and capacity
Management is ramping up capital expenditure, guiding 2026 development CapEx net of partner contributions to $3.25-$3.75 billion, up from $3 billion in 2025. They are investing in a record $10 billion development pipeline, with double-digit yields, and expanding their 5 gigawatt power bank to meet hyperscale demand.
all 6 named companies below
Companiesreturns since call
Supply chain
ANTHROPIC
XAI.PVT
Supply chain
Andy Power
Management references Claude (Anthropic) as one of the leading AI models driving competition, but provides no specific business insight about Anthropic.
Evidence
“the introduction of ChatGPT a few years ago and the ensuing race between Gemini, Claude, Grok, and others marked the beginning of a new chapter”
Supply chain
Andy Power
Hyperscale customers are increasingly making leasing decisions based on who can secure and deliver power capacity on a predictable schedule, favoring operators with visible power banks. — This shifts competitive advantage to landlords with pre-secured power and development track records, potentially raising barriers for new entrants and benefiting incumbents like DLR.
Evidence
“hyperscalers are increasingly making leasing decisions based on who can secure and deliver power capacity on a predictable schedule”
Supply chain
Colin (Head of Hyperscale Partnerships)
Enterprise AI inference demand is emerging in 5-15 MW contiguous blocks, and Digital Realty is seeing early conversations for inference-oriented deployments in its zero-to-one megawatt business. — Inference workloads require proximity to data and networks, potentially accelerating demand for colocation in dense interconnection hubs and benefiting GPU vendors like NVIDIA if enterprise AI scales.
Evidence
“there's emerging conversation around that kind of five megawatt block as inference starts to emerge”
External signals
Supply-chain alpha · 3returns since call
Hyperscale customers are increasingly making leasing decisions based on who can secure and deliver power capacity on a predictable schedule, favoring operators with visible power banks.
Enterprise AI inference demand is emerging in 5-15 MW contiguous blocks, and Digital Realty is seeing early conversations for inference-oriented deployments in its zero-to-one megawatt business.
Digital Realty is taking higher development risk by preparing shells and suites before signing leases, unlike most private capital competitors who wait for a lease to secure financing.
Evidence
“we're probably one of the few in the industry actually taking a little bit more risk in the development of this and getting pad ready... before we have a customer in hand”
Methodology & coverage
Management-only analysis. All 6 validated company mentions are shown. Reported actuals and forward guidance are kept separate. Public evidence is limited to eight short attributed quotes. AI-generated analysis can be incomplete or wrong; verify important claims against the original source.