Why Applied Digital Stock Skyrocketed 38% Last Month But Is Sinking in June

APLD is trading like a leveraged proxy on AI capacity scarcity, but the market is starting to price it as if every announced megawatt becomes immediately monetizable. The real second-order winner is the infrastructure stack around it: grid equipment, switchgear, cooling, and power delivery vendors that get paid on committed builds while APLD still carries execution, funding, and lease-up risk. EKSO’s rename is mostly noise; the important signal is that APLD is separating non-core assets to tighten the equity story around data-center optionality, which can support multiple expansion as long as financing remains available.

The near-term setup is more fragile than the headline momentum suggests. APLD’s equity is effectively duration-sensitive cash-flow optionality, so even a modest repricing in rates can compress the multiple faster than the operating news can re-rate it higher. The current pullback is less about company-specific deterioration and more about the market testing whether AI infrastructure can still command peak multiples if the Fed stays tighter for longer; that makes the next 4-8 weeks a macro beta trade, not a fundamental one.

Contrarian takeaway: the consensus is treating each new data-center announcement as de-risking, but the opposite can happen if capital intensity outruns contracted demand or if power availability becomes the bottleneck rather than the catalyst. If lease-up or financing terms worsen, APLD can trade like a speculative build-out story instead of a scarce AI asset, and that transition usually happens quickly once funding optics shift. On the other hand, if rates stabilize, the stock could snap back sharply because positioning in AI infrastructure remains crowded but not fully hedged.