Where Will Applied Digital Stock Be in 5 Years?

Applied Digital cites a $16.0B contracted lease pipeline, but $11.0B of that depends on a single counterparty, CoreWeave. CoreWeave’s leverage has jumped to ~$21B of total debt (from $7.9B a year earlier), revenue grew ~170% YoY, interest expense now consumes ~25% of sales, and it lost ~$1.2B last year (vs ~$860M prior). OpenAI, a key downstream customer, projects a $14B loss in 2026 and a $115B cumulative cash burn through 2029, meaning funding or demand stress at any link (OpenAI → CoreWeave → Applied Digital) could materially impair Applied’s pipeline and amplify downside given Applied’s own expensive debt.

Analysis

Applied Digital sits atop a one‑tenant-dependent stack where credit and operational risk transmission is nonlinear: a liquidity squeeze at the tenant forces accelerated equipment repossessions, which in turn depresses secondary GPU pricing and extends vendor payment cycles. If repossessions occur, expect a 3–6 month surge in used GPU supply that could knock 20–40% off spot prices and force OEMs to cut near‑term bookings, amplifying margin pressure down the silicon supply chain.

The most actionable horizon is medium term (6–18 months) when refinancing windows and covenant tests cluster. Watch quarterly covenant disclosures, near‑term maturities, and vendor inventory builds — those are the likely catalysts that convert stress into formal distress. Rising high‑yield spreads or a single lender refusing to roll credit will be the fastest path from stress to restructuring.

Second‑order winners include opportunistic buyers of distressed GPU inventories and diversified data center landlords with diversified tenancy and stronger covenants; losers are concentrated build‑to‑suit operators, regional banks with leveraged loans to AI infra players, and any counterparty providing short‑dated financing without recovery rights. A policy/regulatory tightening around asset seizure or equipment recovery could widen recovery loss rates and slow the cyclical clearing of excess hardware.

The consensus is underweighting optionality: Applied’s real estate footprint has option value if the tenant base re‑diversifies or hardware economics normalize, which would revalue long‑dated cash flows. That makes a time‑structured hedge (short near term, retain long‑dated optional protection) a cleaner implementation than an all‑in directional view.

Key Decisions for Investors

• Buy 9–12 month APLD puts (target ~30% OTM) sized 2–3% notional of portfolio to hedge concentrated counterparty risk; max loss = premium paid, upside scenario = 3x–5x if a covenant breach triggers a forced asset sale within 6–12 months.
• Buy protection on CRWV credit: enter 3–5 year CDS or short CRWV bonds/equity (size 1–2%) with a 6–18 month horizon — thesis is widening spreads to 1200–2000bps if refinancing stalls; stop‑loss at 30% adverse move in credit spread basis.
• Long NVDA exposure as a hedge to systemic AI demand (buy 9–12 month calls or LEAPs funded partially by the APLD put premium) — risk: secular GPU weakness; reward: asymmetric upside if final demand remains robust and secondary price dislocations are transient.
• Opportunistic trade: monitor used GPU auction flow — be ready to long selective GPU resale plays or equipment‑leasing names on a 3–9 month basis after a distress event; entry signal = sustained 20%+ slide in spot GPU prices and visibly rising repossession notices.