Cango Inc. Reports Second Quarter 2026 Unaudited Financial Results

Cango (CANG) reported Q2 2026 revenue of $50.8M (down ~50% QoQ) and a net loss of $81.6M, driven largely by $42.9M of non-cash impairment and $8.5M of disposal losses on mining machines. While average cash cost per Bitcoin fell ~5% sequentially to $73,313 and total hashrate reached 27.58 EH/s (19.84 EH/s self-mining plus 7.74 EH/s leased), adjusted EBITDA remained a loss of $10.7M. The company also launched a Bitcoin hedging program and held 1,056 BTC with $31.2M long-term debt as of June 30, 2026, and is onboarding customers for its Georgia AI/GPU build with revenue expected in Q3.

Analysis

The incremental signal is not the smaller loss; it is that CANG is becoming a lower-beta, lower-convexity vehicle. By shrinking hashrate, hedging BTC, and selling marginal machines, management is effectively exchanging upside participation for cash-flow visibility, which should compress any speculative multiple investors were assigning to the treasury/operating leverage story. In the near term that can support the stock if the market rewards "cleaner" financials, but mechanically it reduces sensitivity to a BTC squeeze and makes the equity more of a balance-sheet repair story than a torque trade.

The AI pivot looks optically positive but is still too small to matter at the group level: a few megawatts does not offset a mining base that is being intentionally reduced. That creates a second-order winner/loser dynamic: larger, better-capitalized operators with real power optionality and multi-hundred-MW pipelines (CORZ, IREN, CLSK, MARA) gain relative scarcity value as CANG effectively exits the race for scale. The used-miner disposal also reinforces weakness in secondary ASIC pricing, which is a subtle negative for peers trying to monetize older fleets and for leasing/hosting economics broadly.

Catalyst path: over days, the stock may react well because the headline loss narrows; over 1-3 months, Q3 disclosure on actual AI revenue and BTC hedge marks will decide whether this is a credible diversification or just a small side project; over 6-18 months, the key test is whether the AI platform can reach enough power density and customer utilization to justify a re-rate. What would falsify the bearish view is a sustained BTC move higher without offsetting hedge losses, plus sequential growth in AI hosting revenue that is large enough to offset mining shrinkage. Absent that, the cleaner trade is to fade rallies rather than buy the turnaround.

Key Decisions for Investors

• If CANG rallies on the print, use strength to short/trim into the move rather than chase; frame as a 2-6 week fade if price disconnects from the shrinking operating base.
• Pair trade: long quality miners with real infrastructure optionality (CORZ or IREN) / short CANG for 1-3 months; thesis is that capital will reward scalable AI power exposure, not a 3MW proof-of-concept.
• Buy downside protection on CANG only if borrow is tight: 1-3 month put spreads financed against a post-earnings pop; risk/reward favors limited premium over outright short given name-specific volatility.
• Watch Q3 for AI revenue recognition and hedge accounting detail; if AI contribution is immaterial or hedge losses offset BTC upside, add to the bearish view.
• No actionable view on DGTEF from this release; treat as no-trade until there is evidence of direct financial linkage or supply-chain exposure.