Fifteen-year rent against this year's $249.2 million loss

The length of the contract

The company announced three 15-year leases: about $7.5 billion of contracted revenue for the 300-megawatt Delta Forge 1, about $7.5 billion for the 300-megawatt Polaris Forge 3 and about $5.2 billion for the 210-megawatt Delta Forge 2 signed after the quarter. Those three contracts announced by Applied Digital amount to a revenue commitment spread over fifteen years.^[1]

The numbers standing against that commitment sit in the same release. For the full fiscal year revenue rose 167 percent to $611.3 million, adjusted EBITDA was $107.2 million and the net loss reached $249.2 million. In the fourth quarter, data centre rental revenue consisted of $44.1 million of base rent and $6.5 million of tenant recoveries; the net loss in that same quarter was $110.6 million.^[1]

The timing mismatch

The relationship between those two sets of numbers shows where the credit risk sits. Contracted revenue is a forward promise; construction is present-tense cash. Adjusted EBITDA positive at $107.2 million against a net loss of $249.2 million says the difference gathers in depreciation and financing costs; the items the adjustment leaves out are precisely the items describing how the asset is funded. In my column two days ago I argued that in the debt financing of AI spending the risk concentrates in the most asset-heavy names; the structure here is the next link in that argument, because the collateral is a rent stream that has not yet begun rather than a credit rating. The counter-case is strong too: a fifteen-year contracted lease is the collateral project finance is built on, and if the tenants are strong enough this structure moves closer to an infrastructure asset.^{[1], [2]}

A second move in the same release belongs to the same frame. The cloud services business was separated on May 5, 2026 and merged with Ekso Bionics; the resulting ChronoScale Holdings trades on Nasdaq under CHRN. Applied Digital, however, retains roughly 96 percent of that company. What was separated was the listing; almost all of the economic exposure stays where it was.^[1]

What has to stay true for this not to break

None of this yields a timetable for collapse; what it yields is a threshold that can be watched. Until the rent stream runs ahead of construction spending, the difference is closed with outside funding, and the terms of that funding matter more than the size of today's loss. If the structure is sound, base rent should grow faster than the quarterly net loss; the distance between $44.1 million of base rent and a $110.6 million net loss in the fourth quarter should narrow. If it does not narrow, or widens, what comes into question is the bridge needed to reach that length rather than the length of the contracts.^[1]