Keel Infrastructure: The Demand Was There. The Lease Was Not.

The quarter Keel stopped giving dates

Six months ago Keel Infrastructure was telling investors it would sign 3 leases in 2026. On the third-quarter call, CEO Ben Gagnon retired the target outright: “we no longer expect to sign 3 leases this year” — Benjamin Gagnon, Director and Chief Executive Officer · 2026-11-05. That is the real headline — not the $28 million revenue print, not the $780 million liquidity balance. The company also said it would stop giving lease dates altogether: “we're not going to provide any sort of guidance on lease timings for any of the sites from here on out” — Benjamin Gagnon, Director and Chief Executive Officer · 2026-11-05. For an equity story whose entire value rests on converting 478 megawatts of secured power into contracted cash flow, that is a meaningful negative surprise.

The prior calls had a different tone. In August, Gagnon said: “We're still in active due diligence and negotiations at all 3 of our sites... We remain incredibly optimistic and confident” — Benjamin Gagnon, Director and Chief Executive Officer · 2026-08-10. In March he framed the path more mechanically: “once NTP is clear, that's basically the last thing standing between us and a signed agreement” — Ben Gagnon, Chief Executive Officer and Director · 2026-03-31. The current call moves the goalposts from “when NTP clears” to “when the market and the customer align.”

a lease is not a trophy for a press release. It's a 15-year commitment, and we will not sign the wrong lease at the wrong price just to hit a date.

It is a defensible stance. It is also the same stance Keel has used for several quarters, and the market has now heard it through a 75% drawdown from the June peak. Patience is a strategy until it becomes a schedule problem.

Pennsylvania: the proximate cause

The delay has a name and a place. Governor Shapiro's August executive order made Pennsylvania's grid standards mandatory and took data centers out of the state's fast-track permitting program. Keel says no application has been denied — “not a single permit application has been denied at any of our sites” — Benjamin Gagnon, Director and Chief Executive Officer · 2026-11-05 — but Sharon is down to one final environmental permit in draft, and Panther Creek still has three remaining: one in draft, two in standard technical review. The company also pushed its expansion load-study update from December/January to Q1 2027.

This is not a demand problem; it is a schedule problem. And in a business where the scarce asset is 2027 and 2028 power, schedule is the product. Management spent most of the call arguing that every blocked or delayed project elsewhere makes Keel's remaining sites more valuable. That may be true in theory. But Keel's own permitting and commercial timelines just slipped at the same time, which means the company is not immune to the bottleneck it is marketing.

Demand is not the issue. The tape is.

Gagnon opened with hyperscaler arithmetic: the four largest buyers of data-center capacity are on track to spend more than $800 billion on capex this year; Microsoft, Alphabet and Amazon have a combined contract backlog of roughly $1.8 trillion; Morgan Stanley estimates a 70-gigawatt U.S. data-center power shortfall through 2028. His rhetorical flourish was direct: if the biggest, best-capitalized builders in the world do not have enough 2027 and 2028 power, who does? Keel.

The tape is telling a different story. AI data centers are among the weakest global themes over the last 30 days, with a long list of decliners across semiconductors, power equipment, and neoclouds. Keel's own stock is down 49% in 90 days and 75% from its June peak. The market is not disputing hyperscaler demand. It is discounting the cost and timing of converting that demand into lease revenue — higher rates, wider credit spreads, and a financing market that Gagnon himself described as more selective and more expensive today. The 10-year Treasury hitting its highest level in more than two decades is not a background detail for a company that plans to fund construction with project financing.

The balance sheet is real; the cash burn is real too

The financial engineering is coherent. CFO Jonathan Mir laid out three stage gates: secure power and permits; lock long-lead equipment; only then commit full construction capital against a signed lease and project financing. He said: “We currently do not see the need for any external debt or equity financing in advance of signing a lease.” — Jonathan Mir, Chief Financial Officer · 2026-11-05 Keel has $780 million of unrestricted cash and $380 million of pre-lease commitments through 2028. Moses Lake is fully funded from the balance sheet and expected to be first energized in 2027; Sharon and Panther Creek will require project finance, which is why management is focused on investment-grade tenants or credit wraps. The company has also secured long lead time item equipment for the power connection across all three sites, and is using manufacturing slots rather than full equipment orders to preserve flexibility.

But the cash burn is not theoretical: free cash flow (less SBC) fell to -$106 million, down 59% year over year and 35% sequentially. Revenue fell to $28 million from $68 million as the U.S. Bitcoin mining business was decommissioned; adjusted EBITDA was -$27 million. Keel sold 2,520 Bitcoin for $192 million and now has nominal holdings. The Bitcoin mining business is no longer a meaningful part of the story — only Quebec remains, at 5-6 exahash and a couple of Bitcoin a day. That is a clean strategic pivot, but it also removes the cash engine that used to fund the wait.

What the next catalyst has to be

The bull case is still alive: 478 megawatts across Moses Lake, Sharon and Panther Creek; roughly 2 gigawatts of expansion capacity; multiple potential tenants; no customer has walked away; and a CEO who insists the delay is measured in months, not years. The bear case is simpler: Keel has been “close” on leases for several quarters, the revenue timeline for Pennsylvania now points to early 2028, and the market has stopped paying for narrative. The next catalyst is not another site or another megawatt. It is a signed lease with a financeable counterparty. Until then, Keel is a well-funded developer with the right assets in the right place — and a stock voting on the gap between “right place” and “signed contract.”