Digital Realty Trust, Inc.

▲ Bull Case

• ◆AI demand over-delivers — backlog exceeds $2.0B by FY2027. Additional hyperscaler commitments (AWS, Google) push annual signings toward $2B+, accelerating Core FFO/share CAGR to 14–15%, with FY2029E FFO of $13.50 implying $378/share at 28x P/FFO (+96%).
• ◆PlatformDIGITAL network effects re-rate the multiple. As interconnection approaches $500M (FY2028E) and customers increasingly choose DLR because their partners are already there, the market re-rates DLR from 'hyperscale co-lo' (24–26x) toward 'network platform' (28–30x), matching Equinix's trajectory from 2015–2020.
• ◆Debt refinancing executed at favorable rates. If the $4.9B 2027–2028 maturity wall is refinanced at similar or lower spreads via interest rate normalization, the FFO headwind is eliminated, enabling DLR to hit the upper end of guidance every year through 2029.

▼ Bear Case

• ◆AI capex pause crushes new bookings. If AI monetization disappoints in 2026–2027 and hyperscalers freeze infrastructure spending, quarterly bookings collapse from $350–400M to $100–150M, the backlog depletes without replenishment, revenue growth decelerates to 3–5%, and the multiple compresses from 24x to 18–20x.
• ◆$4.9B debt refinancing at elevated rates. If Treasury yields remain at 5%+ through 2027–2028, DLR refinances $4.9B at +200bp vs. current, creating ~$98M/yr additional interest (~$0.29/share FFO headwind), leading to guidance misses and multiple compression.
• ◆Hyperscaler insourcing accelerates. Microsoft or Google announces large-scale proprietary co-location builds in Northern Virginia or Singapore, competing directly with DLR for their own workloads and creating lease renewal risk and new booking losses from DLR's largest customer concentrations.

“The core debate is 'AI Infrastructure Secular Compounder vs. Cyclical REIT at Peak Demand.' The bullish consensus (32 analysts, avg target $218–220) argues the AI infrastructure supercycle is multi-decade and DLR is one of the two best-positioned pure-plays globally. At 24x P/FFO — below Equinix's 28x — DLR is cheap relative to the infrastructure premium it deserves, and the $1.4B backlog provides 12–18 months of high-visibility de-risked growth. The bearish camp counters that 24x prices in 12%+ FFO/share growth requiring sustained AI capex through 2029 with no pause; the stock has already re-rated +21% in weeks; the $4.9B 2027–2028 debt wall adds leverage overhang; and if AI capex slows, a 15–18x 'old P/FFO multiple' implies 25–35% downside. The debate is not 'is DLR a good business?' (consensus YES) but 'is 24x the right multiple for this point in the cycle?'”

• ◆Q2 2026 earnings (Jul 2026): backlog ≥$1.5B and AI bookings >50% maintained — 70% probability
• ◆Additional hyperscaler announcement (AWS or Google) in DLR markets in Q2–Q4 2026 — 40% probability
• ◆Interconnection revenue >$85M/quarter in Q2–Q3 2026 — 65% probability
• ◆Debt refinancing begins ($4.9B 2027–2028) at spreads at par — 50% probability
• ◆Dividend raise announced for FY2027 (AFFO payout at 65% creates room) — 60% probability
• ◆AI capex slowdown signaled by hyperscalers in Q2–Q3 earnings (negative catalyst) — 20% probability

• ◆AI capex pause: hyperscaler spending freeze collapses bookings to $100–150M/qtr, depleting backlog — 12% probability, −25–35% price impact
• ◆Debt refinancing stress: $4.9B 2027–2028 refinancing at elevated rates creates ~$0.29/share FFO headwind and earnings misses — 10% probability, −10–15% FFO impact
• ◆Hyperscaler insourcing >$500M in contracted value: Microsoft/Google proprietary builds threaten renewal and new bookings — 8% probability, −15–20% price impact
• ◆Interest rate spike (10-yr >5.5%): multiple compression on REIT valuations — 15% probability, −15–25% price impact
• ◆New supply in constrained markets reducing pricing power — 10% probability, −5–10% price impact