Cango (CANG) Q3 2025 Earnings Call Transcript

Cango reported Q3 revenue of $224.6M, up 60.6% sequentially, driven by Bitcoin mining revenue of $220.9M and production of 1,930.8 BTC; net income was $37.3M (versus a prior-year loss) and adjusted EBITDA was $80.1M. Deployed hashrate is 50 EH (operating ~46 EH with ~90% uptime); average mining cost was $81,072/ BTC excl. depreciation and $99,383 all-in. Balance sheet items: $44.9M cash, $660M bitcoin collateral receivables, $365.7M net mining machines, and $405.1M long-term related-party debt; the company acquired a 50 MW Georgia facility and is pursuing clean-energy projects in Oman and Indonesia to support a phased move into distributed AI compute. Management reaffirmed a mine-and-hold reserve policy and emphasized risk controls, but Bitcoin price volatility and power curtailments remain primary downside risks.

Analysis

Cango’s strategic pivot creates an asymmetric optionality profile: the asset-light footprint preserves capital agility while selective ownership of energy assets can unlock meaningful per-unit margin capture if power pricing and offtake terms are favorable. That hybrid model shifts the battleground from pure scale to power-cost control and operational flexibility, forcing large, highly leveraged miners to compete on different terms (price vs. controllable fixed-cost exposure).

The biggest underpriced risks are balance-sheet concentration and financing optionality under stress. Pledged or encumbered crypto reserves and related‑party funding layers raise the likelihood of forced deleveraging if Bitcoin prices retrace sharply or if counterparties accelerate maturities — a liquidity shock would compress valuation multiple far faster than an equivalent drop in coin price due to covenant mechanics.

Execution on the AI compute pivot is a multi-dimensional, multi-year play: margin realization depends on (1) securing low-cost, dispatchable green power, (2) proving a distributed orchestration stack that wins mid-market customers, and (3) avoiding heavy capex commitments that recreate hyperscaler economics. Each leg is a discrete catalyst: wins will re-rate optionality; delays will expose the company to reversion toward commodity-miner multiples.

Monitor four high-signal datapoints: realized BTC monetization behavior (sell vs hold cadence), fraction of unencumbered reserve collateral, time-to-commercialization for green projects, and contract wins in distributed AI pilots. Near-term upside is real but paired balance-sheet and execution risks create sharp binary outcomes.

Key Decisions for Investors

• Long CANG equity (size 2-4% of crypto/infra sleeve) with a 6–12 month horizon; pair with a protective tail via BTC put exposure (or short BTC futures) sized to cap portfolio downside to ~20–30%. Rationale: capture re-rating from successful hashrate optimization and early AI pilot traction while limiting systemic crypto drawdown risk.
• Buy a calendar/options spread on CANG: long-dated call (12 months) financed by selling near-term calls (1–3 months). Time this after any near-term run-up to exploit higher short-term implied volatility and retain long optionality on the AI pivot. Reward: asymmetric upside if multi-quarter execution continues; risk: premium paid if execution stalls.
• Relative-value pair: long CANG / short a large, highly levered miner (e.g., MARA/RIOT) sized 1:1 notional for a 3–9 month trade. Thesis: Cango’s asset-light + selective asset ownership benefits from margin arbitrage and operational agility; short leg hedges systemic BTC downside. Monitor financing spreads and uptime metrics as triggers.
• Maintain a small, permanent portfolio tail-hedge via liquid inverse-BTC instruments or deep OTM BTC puts (ongoing). This insurance is inexpensive relative to the asymmetric hit from a forced deleveraging event tied to pledged collateral or covenant breaches.