Is Applied Digital Stock Going to $0?

Applied Digital (APLD) has rapidly reoriented to AI-focused data-center leasing, growing revenue from $55 million in 2023 to $264 million over the last four reported quarters and securing up to $16 billion of lease commitments over the next 15 years. However, the company lost $125 million in the last 12 months and has ballooned debt from ~$42 million in Q1 2024 to nearly $2.6 billion as of Nov. 30, largely financed at ~9.25% interest, while its future cash flows are highly concentrated in a single tenant, CoreWeave, which itself is loss-making and highly levered—creating significant counterparty and leverage risk for creditors and equityholders.

Analysis

Market structure: Applied Digital (APLD) sits at the intersection of scarce AI-capable data-center supply and hyper-concentrated demand — it has >$16B of lease commitments over 15 years and revenue surged from $55M (2023) to $264M LTM. Winners are AI hyperscalers and chip makers (NVDA, INTC) that keep server demand growing; losers are levered, single-tenant landlords if tenant credit deteriorates (primary risk: CRWV exposure). Limited specialist supply gives pricing power but counterparty concentration transfers most credit/operational risk to landlords.

Risk assessment: The balance sheet moved from $42M to ~$2.6B debt by Nov 30 with ~9.25% financing — refinancing and interest-rate tail risks are material. Short-term (days–months) volatility will center on CoreWeave liquidity events and covenant triggers; medium-term (6–18 months) risks include missed lease payments or higher cost of capital; long-term (>18 months) execution risk around diversification and path to positive FCF. Hidden dependency: APLD’s equity value is functionally a credit on CRWV’s solvency and grid/power access, not pure AI demand.

Trade implications: Favor asymmetric hedges: small, option-backed short exposure to APLD and long exposure to NVDA/large diversified REITs. Cross-asset: expect credit spreads on specialty infra names to widen; HY paper in the space will underperform investment-grade datacenter bonds. Timing: act within next 30–90 days ahead of likely refinancing/earnings windows.

Contrarian angle: The market may over-penalize APLD if leases are non-recourse and include escalators; a confirmed lease diversification or a >50% reduction in CRWV revenue share would be a binary positive. Watch for M&A interest from large REITs or covenant amendments that could rerate equity; until then downside asymmetry dominates.

Key Decisions for Investors

• Establish a 1–2% net short position in APLD (ticker APLD) sized by portfolio, implemented as a 12–18 month put spread to cap premium (buy put ~30–40% OTM, sell deeper OTM) to profit if stock re-rates on refinancing/counterparty stress; act within 30 days ahead of year-end filings.
• Establish a 2–4% long position in NVIDIA (NVDA) as a directional play on sustained AI hardware demand over 6–24 months; use covered-call overlays if volatility rises above historical 60–90 day realized vol to harvest premium.
• Initiate a pair trade: long 2% position in a diversified data-center REIT (e.g., Digital Realty or equivalent) and short 1% APLD to capture idiosyncratic landlord credit risk while keeping exposure to secular AI infra.
• Monitor three binary triggers over the next 60–120 days before changing stance: (1) CoreWeave liquidity event or missed payment, (2) APLD debt refinancing terms >10% coupon or material covenant waivers, (3) reduction of CoreWeave’s share of APLD contracted revenue below 50% — if any occur, increase APLD short to 3–4% and buy additional puts.