Cango sells 4,451 BTC for $305M to cut leverage and bankroll AI GPU rollout

Over a single weekend Cango converted 4,451 BTC into approximately $305 million denominated in USDT and applied the net proceeds to extinguish a bitcoin‑backed loan, a move management described as targeted deleveraging to improve balance‑sheet flexibility. The firm signaled a strategic reorientation: it intends to repurpose portions of its power‑anchored, grid‑connected campuses to host containerized, modular GPU clusters focused initially on inference workloads for small and mid‑sized enterprises while preserving bitcoin‑mining operations. Cango also appointed a CTO for the AI business with experience in large‑scale GPU orchestration, underscoring an intent to operationalize rather than merely signal an AI optionality. The asset sale reduces immediate refinancing and margin‑call risk by trimming BTC collateral exposure and creating liquidity for capex and operations. That capital allocation decision echoes a broader sector trend: miners facing halving‑driven revenue pressure are converting stranded power and real estate into colocation and AI compute capacity to capture emerging demand. Market signals have supported this window of opportunity—short‑term easing in network hashrate and mining difficulty, alongside a modest uptick in hashprice, have widened the economic case for non‑SHA‑256 workloads at some sites. Concurrently, hyperscalers’ stepped‑up AI capex and multiyear capacity commitments have created a near‑term market for ready‑to‑energize megawatts, shortening deployment timelines for enterprise customers. However, the commercial upside is bounded by practical constraints: accelerator supply shortages, OEM backlogs, packaging and test throughput, and local permitting or community pushback can materially slow deployments. The economics also favor sites with firmed, low‑carbon power as hyperscalers increasingly demand guaranteed 24/7 supply, raising the premium on dispatchable renewable plus battery projects. If Cango can deploy modular GPU capacity quickly, secure customer agreements, and achieve predictable utilization, the pivot could produce steadier, contract‑driven revenue and reframe investor comparables away from pure BTC exposure. Failure to overcome supply‑chain bottlenecks, secure firmed power, or build software and sales channels would leave the company with fewer BTC reserves and the same operational risks it faced before. In short, the BTC sale buys time and optionality—reducing near‑term leverage while funding a bet on higher‑margin, non‑BTC workloads whose payoff depends on execution, market demand, and power economics.