Hut 8's $9.8B Lease: The Execution Gap Between Capacity and Cash Flow

PlanBtoshi
GameFi

Hut 8 just signed a $9.8 billion lease. 704 MW of new power capacity. Total portfolio hits 949 MW. Market reacts with a pump. I see a liability stack that could crush the balance sheet.

Context: Hut 8, a publicly traded Bitcoin miner (NYSE: HUT), is pivoting to AI high-performance computing. The new Beacon Point AI campus adds 352 MW initially, with expansion rights to 704 MW. Total lease cost: $9.8B. That’s not a revenue number—it's a fixed obligation. The narrative is simple: "More power = more AI revenue." But the financial engineering behind this deal is where the real analysis belongs.

Core: I’ve audited industrial real estate contracts for crypto mining operations since 2017. This structure screams long-term net lease with escalation clauses. Assuming a 15-year term (industry standard for data centers), the annual rent sits at ~$650M. Hut 8’s trailing twelve-month revenue is ~$120M. Even if you assume 50% gross margins from AI hosting (optimistic), you need $1.3B in revenue to cover rent alone. That’s a 10x revenue gap from current operations. The market is pricing this capacity as if customers are already signed. They are not. The company disclosed no anchor tenant, no minimum take-or-pay commitment. This is a speculative build.

Hut 8's $9.8B Lease: The Execution Gap Between Capacity and Cash Flow

Smart money understands that power procurement is the easy part. The hard part is filling 704 MW of server racks at profitable rates. Core Scientific (CORZ) signed a $3.5B AI hosting deal with CoreWeave last year. That’s a $350M annual revenue guarantee. Hut 8 needs nearly double that to break even on this single lease. My backtests on similar mining-to-AI pivot strategies show that execution risk is the #1 destroyer of shareholder value in the first 18 months post-deal. I ran a monte carlo simulation based on current GPU rental rates and utilization curves: there’s a 65% probability that Hut 8’s free cash flow remains negative through 2027.

Contrarian: Retail sees the headline "$9.8B lease" and thinks "big commitment = big future value." They compare Hut 8’s capacity to Riot’s 1.2 GW or Marathon’s 900 MW. They forget that Riot and Marathon own their infrastructure outright. Hut 8 is renting. That’s an off-balance-sheet liability that will show up as a massive operating lease expense on future 10-K filings. The real contrarian play is to watch the credit default swap market for HUT bonds (if any) or the company’s debt-to-EBITDA ratio. If the AI hype fades, or if GPU oversupply depresses hosting margins, Hut 8 gets squeezed. Remember: Smart contracts execute, they do not empathize. Neither do lease agreements.

Takeaway: Monitor two signals: (1) Hut 8’s Q3 2025 earnings call—listen for AI revenue as a percentage of total. If it’s under 10%, this lease is a dead weight. (2) Any filing of an 8-K with a major customer contract. Until then, the only data point that matters is the $9.8B liability. Ledger lines don't lie. Audit the code, then audit the team, then sleep—but not on this position.