Keel Infrastructure (KEEL) Q3 2026 Earnings Call Transcript

Keel Infrastructure's Q3 2026 revenue fell to $28 million from $68 million a year earlier as it wound down U.S. Bitcoin mining; continuing-operations loss was $62 million, including a $55 million noncash capped-call loss, and adjusted EBITDA was negative $27 million. Management reported approximately $780 million of unrestricted cash and 478 megawatts of near-term secured capacity, but withdrew its goal of signing three leases by year-end after Pennsylvania permitting and commercial timelines slipped by months. Moses Lake is under construction and expected to be energized in 2027; higher interest rates and more selective debt markets add financing risk to the development pipeline.

Analysis

KEEL is now best viewed as a permitting-and-financing option, not an operating AI infrastructure compounder: legacy mining cash generation has largely disappeared, while meaningful HPC revenue depends on leases, construction and energization. The reported liquidity provides runway, but it does not eliminate the key risk—higher financing costs or weak tenant credit could make projects uneconomic even if power is available. The $55 million capped-call mark is noncash; the more important signal is negative adjusted EBITDA alongside elevated development commitments.

The absence of a lease matters more than management’s characterization of tenant interest. Pennsylvania permitting delays have already broken the year-end leasing target, and the company has declined to give a replacement date. That leaves near-term valuation exposed to milestone slippage and makes claimed scarcity value difficult to underwrite until a binding lease and project financing establish contracted yield. Conversely, keeping long-lead equipment flexible and tying major construction spend to leases limits speculative-capex risk.

Over 1–3 months, watch for a Moses Lake lease, final Sharon/Panther Creek permits, and Q1 2027 load-study results. Over 6–18 months, the thesis requires on-time energization, financeable tenants and project-level returns that absorb prevailing rates. Falsifiers: further permit slippage, a lease without disclosed credit support/economics, financing terms that weaken returns, or liquidity falling faster than commitments. Hyperscalers’ capacity needs support third-party demand, but could also raise their deployment costs; this is not automatically a margin-positive outcome for AMZN, GOOG, META or MSFT.

Key Decisions for Investors

• Keep KEEL on a milestone-driven watchlist rather than buying the demand narrative alone. Reassess on a signed lease with tenant credit support, RFS timing and project-financing terms disclosed; without those, avoid sizing it as a proven HPC landlord.
• For existing KEEL exposure, underweight or hedge into strength if permitting slips again or liquidity declines materially against the stated development commitments. Do not treat the capped-call accounting loss as the core risk; focus on cash burn and financing economics.
• No broad trade in AMZN, GOOG, META, MSFT or NVDA from this call alone: capacity scarcity may support demand for infrastructure while simultaneously increasing customers’ cost to deploy. Track hyperscaler capex and third-party capacity commentary for confirmation.
• Verify the next permit decisions, Moses Lake lease/financing package, and the Q1 2027 Panther Creek/Scrubgrass load-study update. These are the catalysts that can convert KEEL’s pipeline into bankable assets—or expose the gap between potential megawatts and deliverable capacity.