This Artificial Intelligence (AI) Infrastructure Stock Has Been Crushing Nvidia on the Market. Its Latest Deal Could Send the Stock Even Higher

Applied Digital has signed a new 15-year AI factory lease worth $5.2 billion with a U.S.-based hyperscaler, potentially rising to $12.7 billion over 30 years if renewals are exercised. The company now has contracts for five AI factory campuses and a reported lifetime lease revenue base case of $36 billion, with pipeline potential up to $86 billion. Fiscal 2026 revenue is estimated to have jumped 96% to $422 million, though the stock remains expensive at 35x sales.

Analysis

APLD is increasingly looking less like a pure "AI hype" equity and more like a contracted capacity arb: the market is beginning to underwrite long-duration cash flows before the assets are fully built. That creates a second-order effect where financing terms, execution cadence, and power procurement matter more than near-term revenue prints; each incremental campus that gets financed and delivered de-risks the next one through a lower cost of capital and better customer credibility. The main beneficiary is APLD's own equity, but the broader winner is any supplier set tied to power, cooling, and electrical infrastructure because hyperscaler demand is now translating into multi-year buildouts rather than short-cycle orders.

The risk is that valuation is now being priced off a very long runway while the company is still in the heavy capital-expenditure phase. If build timing slips, interconnect delays emerge, or a customer pauses expansion, the equity can re-rate quickly because the multiple is effectively underwriting future lease monetization, not current earnings power. Over the next 3-12 months, the key catalyst is evidence that new capacity is turning into contracted, financeable project delivery faster than the burn rate expands; if that cadence breaks, the stock's premium can compress sharply even if the long-term thesis remains intact.

The market is likely underestimating how much this model benefits incumbent hyperscalers and neoclouds by outsourcing infrastructure risk, while simultaneously pressuring smaller peers that lack balance-sheet scale or land/power access. There is also a supply-chain knock-on: persistent AI data center demand should keep pricing firm for transformers, switchgear, generators, and liquid-cooling systems, creating a broader infrastructure trade beyond semis. The contrarian view is that the move may be over-owned as a "pure AI beneficiary," when in reality the key variable is project finance execution; in a risk-off tape, long-duration infrastructure names with high sales multiples tend to de-rate before the cash flow story fully matures.

Key Decisions for Investors

• Long APLD on pullbacks over the next 2-6 weeks, but size modestly; use a 10-15% trailing stop because the equity is now primarily a financing/execution story, not a self-funding utility.
• Pair trade: long APLD / short a weaker, non-contracted AI infrastructure peer with less power access and lower visibility over 3-6 months; the spread should benefit from contract-backed revenue visibility and better capital access.
• Buy call spreads on APLD with 3-6 month tenor to express upside while capping multiple-compression risk; this is cleaner than outright stock given the high sales multiple and event-driven volatility.
• Add a basket long in electrical infrastructure beneficiaries (e.g., ETN, HWM, VRT) as a second-order AI trade; this should outperform if data center buildouts remain the dominant capex theme over the next 6-12 months.
• Avoid chasing NVDA on this specific headline; the incremental read-through is more about downstream power and data-center ecosystem winners than about upside to the core AI chip leader.