Could Applied Digital Stock Help You Become a Millionaire?

Applied Digital is pivoting from crypto to AI-focused hyperscale data centers, building Polaris Forge 1 (450 MW) and Polaris Forge 2 (300 MW) with completion expected in 2027 and beginning Delta Forge 1 (initial 430 MW) potentially mid‑2027. The company has signed lease deals totaling $11 billion with CoreWeave and roughly $5 billion with another unnamed AI customer (both over 15 years), while Wall Street estimates revenue of $346 million for fiscal 2026 and $535 million for fiscal 2027; the stock has risen ~700% over 18 months to a $9.5 billion market cap and currently trades at about 17–18x next‑year revenue. Despite large long‑term contracts and capacity expansion, the firm is burning cash to fund construction, relies on a small number of large customers, and faces execution and demand risk if AI investment momentum slows.

Analysis

Market structure: APLD (Applied Digital) and its anchor customers (CoreWeave/CRWV and the unnamed AI lessee) are short-term winners as APLD brings ~1,180 MW (450+300+430 MW) online by 2027, capturing premium scarcity rents for specialized AI colocation. Traditional generalist colo REITs (e.g., DLR) and smaller hyperscalers could lose pricing power where latency/power density matter; expect localized power/permitting bottlenecks to sustain pricing through 2027 but compress thereafter if many entrants scale simultaneously.

Risk assessment: Key tail risks are customer-concentration (CoreWeave ~$11bn + unnamed ~$5bn = >$16bn revenue backlog), financing/dilution risk while burning cash to 2027, grid/permit curtailments, and an AI-capex pullback. Immediate (days–weeks): mean reversion/volume-driven pullbacks; short-term (3–12 months): refinancing and milestone delivery risk; long-term (2027+) depends on lease commencements and GPU demand sustaining utilization above ~70% to hit modeled revenues.

Trade implications: Favor operational contractors and GPU vendors (CRWV, NVDA) over speculative campus developers: CRWV benefits from secured demand; NVDA benefits from sustained chip scarcity. Practical strategies: small long allocations to CRWV, protective hedges on APLD, and underweight generic REITs into targeted AI infra/semiconductor exposure; watch power/energy inputs (electricity cost moves >20% materially change IRRs).

Contrarian angles: Consensus overlooks execution/delivery timing and energy grid risk—APLD’s 700% run presumes flawless on‑time builds and non-dilutive financing. Historical parallel: crypto-data-center cycle (2017–19) where operator economics reversed when hardware/pricing moved; if GPU spot prices drop >20% or a major lessee delays, APLD's revenue visibility and valuation (17–18x FY27 revenue) look vulnerable to 30–60% downside.

Key Decisions for Investors

• Establish a tactical 2–3% long position in CRWV (CoreWeave) over 6–18 months to capture operator upside from signed leases; take profits at +40% and cut at -15%.
• Implement a relative-value pair trade: short APLD equal to 1–2% of portfolio notional while long CRWV at 2–3% (dollar-neutral) with a 6–12 month horizon; target spread capture of 30–50%, stop-loss if APLD rises +25% vs entry.
• If holding/adding APLD exposure, buy a 3–6 month put-spread protection roughly 10–20% OTM sized to cover 50% of position cost (cost target <2% portfolio) to guard against execution/financing surprises ahead of 2027 build milestones.
• Reduce generic data-center REIT exposure (e.g., DLR) by 3–5% and redeploy into NVDA via 9–12 month calls or direct stock exposure on >10% dips; reduce if GPU spot price declines >20% within 30–60 days or if APLD/CRWV revenue recognition slips by >90 days.