Stock-Split Watch: Is Applied Digital Next?

Applied Digital, which rebranded from Applied Blockchain in Nov 2022, reported Q2 FY2026 revenue of $126.6 million, up 250% year-over-year, and carries roughly $16 billion of backlog from two hyperscaler leases (including $11 billion tied to CoreWeave and a $5 billion, 15-year lease with an unnamed U.S. hyperscaler). The company has seen its stock surge (~1,720% over three years as of Jan. 12) but remains unprofitable, trading at ~33x trailing sales while incurring heavy capex (from $141.8M in FY2024 to $681.6M in FY2025 and $801.5M in the first two quarters of FY2026), and faces customer-concentration risk despite advantaged low-energy, high-density data-center assets.

Analysis

Market structure: Winners are AI hyperscalers (CoreWeave CRWV) and GPU suppliers (NVDA) plus specialist builders like Applied Digital (APLD) that lock long leases; losers are legacy colocation REITs (e.g., EQIX, DLR) and small data-center operators facing pricing pressure. APLD’s $16B backlog vs ~$506M annualized revenue implies ~32x coverage, signaling multi-year forward demand but extreme concentration risk. Cross-asset: expect higher commodity (power/natural gas) local prices and elevated equity implied vol for APLD; corporate credit spreads could widen if APLD funds capex with debt or equity issuance.

Risk assessment: Tail risks include a hyperscaler contract renegotiation/default, grid/power shocks raising operating costs >20%, or GPU shortages delaying revenue ramp—each could wipe 30–50% of projected free-cash-flow. Near-term (days–weeks): earnings/lease KPIs and any equity-debt announcements; medium (3–12 months): capex funding events and first MW revenue recognition; long-term (2–5 years): secular AI compute demand vs stranded-capacity risk. Hidden dependencies: firm power purchase agreements, permitting timelines, and GPU procurement cadence.

Trade implications: Direct: establish a tactical 2–3% long APLD position (6–12 month horizon) sized to tolerate 40% drawdown, and buy a 9-month 25% OTM put for downside protection. Pair trade: long APLD (2%) / short EQIX or DLR (1–1.5%) to express re-rating of AI-optimized infra vs legacy colo. Options: consider a 6-month call spread on APLD to limit premium outlay if positive catalysts (lease milestones) hit.

Contrarian angles: The market underestimates execution and concentration risk—backlog is large but recognition timing and margins are uncertain; parallels to 2017 crypto buildouts show high risk of stranded capacity if demand growth slows. Watch for share issuance, GPU supply notices, and signed power contracts—each is a binary catalyst that can move the stock 30–50% within 90 days.

Key Decisions for Investors

• Initiate a 2–3% portfolio long position in APLD (Applied Digital) with a 6–12 month horizon; simultaneously buy a 9-month put 25% below purchase price to cap downside to ~15–20% net cost.
• Construct a relative-value pair: long APLD (2%) vs short EQIX or DLR (1–1.5%) to isolate re-rating toward AI-optimized high-power campuses; rebalance if APLD outperforms by >30% or underperforms by >20%.
• Buy a 6-month APLD call spread (buy nearer-term ATM call, sell OTM call ~30% above) sized to 1% of portfolio to leverage positive lease/operational milestones while capping premium.
• Reduce exposure to legacy colocation REITs by 2–4% over the next 3 months; rotate proceeds into AI infra suppliers (NVDA) and regional power plays only if power contracts/PPA terms confirm long-term economics.
• Monitor three binary triggers over the next 90 days before increasing size: (1) disclosure of customer contract cancellability/termination clauses, (2) announced equity or debt raise size/timing, (3) signed long-term power purchase agreements covering >50% of planned MW.