1 Incredible AI Stock that Wall Street Thinks Can Soar 59% Over The Next Year

Applied Digital raised $2.35 billion in Q2 at a 9.25% rate and its stock trades near $28.50 versus an average analyst target of $45.27 (~60% upside). The company is building multiple contracted AI-focused data centers (Polaris Forge 1: 100MW active, expanding to 400MW; Polaris Forge 2: +200MW starting construction) to capture elevated AI demand. Key risk is an aggressive, debt-financed buildout with expensive financing despite 15-year leases, prompting the author to advise caution and suggest lower-risk AI investment alternatives.

Analysis

Winners are likely to be pure-play GPU hardware and software stack providers that avoid heavy real-estate capital intensity: they capture margin expansion from AI compute without the project-level refinancing risk. A second-order beneficiary is counterparty credit providers (utilities, PPAs, equipment vendors) that can extract stronger collateral terms or accelerated payments as developers push to de-risk buildouts. Conversely, highly levered operators are exposed to a liquidity reflexivity: a single delayed ramp or tenant bankruptcy propagates through contractor liens, equipment repossession timelines, and sponsor equity wipeouts.

Key risks cluster around financing and timing rather than demand per se. If wholesale funding conditions retrace towards wider spreads or if supply-side GPU availability relaxes (reducing urgency for third-party capacity), the valuation multiple on operating cashflows can compress sharply within 6–18 months. Watch milestone-driven windows — upcoming quarterly updates or bond/covenant test dates can produce outsized moves on relatively small slippage in construction or leasing cadence.

From a competitive-dynamics angle, tenant concentration and single-source cloud partnerships create asymmetric counterparty risk: a large customer negotiating better on-rates or stepping away during a downturn can flip economics overnight. That same concentration creates an M&A option for strategic buyers (hyperscalers, chip vendors) who value immediate scale; this is a non-linear upside that consensus tends to underprice but is contingent on regulatory and financing appetite. Monitoring PPA terms, contractor lien filings, and GPU allocation announcements will give early signals of either stress or takeout interest.

Key Decisions for Investors

• Short APLD equity (or buy 12-month OTM puts) sized 2–3% NAV. Thesis: refinancing/timing slip triggers >30–50% downside within 6–12 months. Protect with a 20–25% stop or hedge with short-dated calls to limit premium risk.
• Pair trade: Short APLD / Long CRWV equal notional for 6–12 months. Rationale: capture balance-sheet discount vs operating GPU cloud exposure; target relative spread compression of 20–40% in favor of CRWV. Risk: systemic re-rating of AI infrastructure lifts both — cut losses if both move >15% higher in 4 weeks.
• Long NVDA exposure via 9–15 month calls (or outright equity) as a lower-capital-intensity way to access AI compute demand. Position sizing 2–4% NAV; target asymmetric 2x+ upside if secular GPU tightness persists, max loss = premium paid.
• Credit hedge: Buy protection on APLD paper (3-year CDS) or short senior unsecured bonds. If credit spreads widen by 250–400bp, expect bond prices to fall ~20–35%; size to offset equity short exposure and limit tail risk from systemic dislocations.