Applied Digital secures $7.5B hyperscaler lease for AI campus

Applied Digital signed a 15-year lease with a U.S.-based hyperscaler for 300 MW at its Delta Forge 1 campus, representing about $7.5 billion in contracted value. Total contracted lease revenue now exceeds $23 billion, with more than 50% backed by investment-grade customers, and the company also plans up to $600 million in new bridge and revolving credit facilities to fund development. Shares were already up 695% over the past year, and analysts remain bullish with price targets of $40 to $45.

Analysis

APLD’s win is not just revenue visibility; it is an option on scarce power conversion. The market is increasingly treating contracted megawatts as a proxy for future cash flow, but the more important second-order effect is financing power: every added long-dated, investment-grade tenant lowers perceived execution risk and should compress project-level funding spreads, which matters more than headline lease value for a capital-intensive developer.

The real competitive edge is not the hyperscaler relationship itself, but the ability to keep stacking financed projects before the market questions build-out discipline. If APLD can turn these announcements into a repeatable underwriting model, smaller private data-center developers and non-IG peers will struggle to match its access to bank capital, even if they can source land and interconnects. That said, this also creates a reflexive setup: the equity is now partly priced as a financing vehicle, so any delay in permits, grid hookups, or customer ramp could hit the stock harder than a normal lease miss.

Contrarian takeaway: consensus is underweighting duration risk. A 15-year contract sounds de-risked, but the equity is exposed to a 2-3 year construction window where inflation, equipment lead times, and power-availability slippage can erode project IRRs before revenue starts. The stock has already repriced for success; the next leg likely depends on whether management can execute without constant dilution, not on whether demand exists.

EKSO remains a side-show unless the proposed transaction becomes more material, but the presence of linked equity awards suggests some optionality around corporate activity. For APLD, the cleaner catalyst path is credit-market validation and additional hyperscaler closes over the next 3-9 months; absent that, the shares can mean-revert sharply if investors start valuing the backlog at a discount to execution risk.

Key Decisions for Investors

• Maintain a tactical long APLD, but use a 6-9 month horizon and size it as a financing/execution story rather than a pure AI demand trade; upside is another validation close or cheaper project debt, while downside is a sharp re-rate if build timelines slip.
• Buy APLD upside via call spreads rather than outright stock into the next 1-2 quarters: the thesis is that additional hyperscaler announcements and debt facility close-outs can extend momentum, but implied volatility and execution risk make defined-risk exposure superior.
• Pair long APLD / short a weaker, non-investment-grade data-center developer or broader speculative AI infra basket over 3-6 months; APLD’s bankability and IG tenant mix should outperform if capital markets remain selective.
• Set a tight risk trigger on APLD if management announces any dilution, covenant stress, or project delay beyond the current mid-2027 start window; that would likely compress the multiple quickly because the stock is already pricing in smooth execution.
• Do not chase EKSO on this headline alone; treat it as a corporate-action optionality name only if the proposed transaction gains concrete terms, otherwise the signal-to-noise is poor.