Bitcoin mining in China rebounds, defying 2021 ban

Bitcoin mining activity is resurging in China despite the 2021 ban, with Hashrate Index estimating a 14% national share of global mining at end-October and CryptoQuant estimating 15–20% of global capacity now operating in China. Factors cited include cheap regional electricity (notably Xinjiang and Sichuan), excess data-center capacity, and higher bitcoin prices; rig-maker Canaan reported China accounted for 30.3% of its global revenues last year (vs. 2.8% in 2022) and, per a source, over 50% of sales in Q2. The rebound could provide demand support for bitcoin and lift mining-equipment vendors, while signalling potential gradual easing or pragmatic enforcement of Chinese policy toward digital assets.

Analysis

Market structure: The on‑shore resurgence reallocates ~15–20% of global hashing to lower‑cost Chinese clusters, creating immediate winners among ASIC OEMs (CAN) and regional hosting/data‑center service providers while pressuring higher‑cost North American miners’ margins. Expect global difficulty to rise ~10–20% over 1–3 months as new capacity stabilizes, mechanically reducing BTC revenue per TH and compressing EBITDA for legacy, high‑opex rigs. Secondary markets for used ASICs should firm near term, supporting equipment manufacturers’ gross margins and elevating pricing power for scarce new models.

Risk assessment: Key tail risks are renewed provincial or national enforcement (20–30% within 12 months), grid curtailments in Xinjiang/Sichuan tied to seasonal dispatch, and a potential flood of resold ASICs that collapses ASPs. Immediate (days) volatility will track policy headlines and CAN sales prints; short term (weeks–months) driver is inventory digestion and difficulty adjustment; long term (quarters) the structural outcome hinges on Beijing’s pragmatic tolerance vs. formal relaxation. Hidden dependencies: logistics, financing for Chinese miners, and thermal fuel prices (coal/gas) can swing break‑even electricity costs ±30%.

Trade implications: Tactical alpha is in equipment OEM exposure and relative miner dispersion. Prefer option structures to express demand upside while limiting regulatory tail risk: buy CAN call spreads into the next two earnings; short high‑cost US miners (MARA/RIOT) to capture margin divergence; rotate modest crypto beta into spot BTC or 6‑month calls if difficulty materially outpaces price. Time entries on CAN on a <=10% pullback or after confirmation of sustained sales in the next 30–45 days; trim at +30–50% or on adverse policy signals.

Contrarian angles: Consensus underestimates durability of policy risk and overestimates immediate demand sustainability — a temporary buying binge by resellers can masquerade as structural recovery. Historical parallel: 2019–2021 episodic miner migrations created transient demand spikes then capital flight; if ASIC ASPs fall >25% in 60 days, OEM revenue reversion becomes likely. Watch ASIC resale prices, provincial dispatch notices, and CAN region‑level revenue disclosure as high‑signal leading indicators.

Key Decisions for Investors

• Establish a 2–3% NAV long position in Canaan (CAN) via a 3‑month call spread: buy 15% OTM calls and sell 40% OTM calls sized to equal 2% NAV, enter within 30–45 days or on a <=10% share pullback; target exit +30–50% or on a provincial enforcement notice.
• Initiate a 1–2% NAV short position in high‑cost US miners (e.g., MARA, RIOT) to capture margin divergence; average in over 2–6 weeks, place stop‑loss to cover if BTC rallies >25% from current levels or miner margins improve.
• Deploy 1% NAV to directional BTC via 6‑month calls ~20% OTM (or equivalent spot ETF exposure) to play demand support from Chinese hashing; hedge with a 10% trailing stop or sell into a 40% BTC rally.
• Monitor three high‑signal triggers for position changes over the next 30–60 days: (1) CAN regional revenue disclosure; (2) provincial grid/energy dispatch orders in Xinjiang/Sichuan; (3) secondary market ASIC ASPs — cut CAN to zero and cover miner shorts if any show formal nationwide enforcement language or ASIC ASPs drop >25% in 60 days.