Public Bitcoin miners sold 32,000 BTC in Q1 2026, exceeding total 2025 liquidations and the previous quarterly record of 20,000 BTC set during the Terra-Luna collapse in Q2 2022. The sell-off coincided with hashprice falling to $28–30 per PH/s/day, the lowest reading since 2018, and Bitcoin tradi...
"We're not a Bitcoin mining company anymore. We're an energy infrastructure company that happens to mine Bitcoin." — Asher Genoot, CEO of Hut 8, Yahoo Finance interview, April 2026
Public Bitcoin miners sold 32,000 BTC in Q1 2026, exceeding total 2025 liquidations and the previous quarterly record of 20,000 BTC set during the Terra-Luna collapse in Q2 2022. The sell-off coincided with hashprice falling to $28–30 per PH/s/day, the lowest reading since 2018, and Bitcoin trading below $60,000 for portions of the quarter.
The data tells a split-screen story. Pure-play miners face margin compression: the weighted-average cash cost to produce one BTC rose to approximately $80,000 by Q4 2025, according to CoinShares, while Bitcoin traded at $78,178 as of May 1, 2026. Meanwhile, miners that pivoted infrastructure toward AI and high-performance computing (HPC) hosting have collectively signed over $70 billion in long-term contracts. Their stocks outperformed Bitcoin by 70% year-to-date.
The $77.6 billion public mining sector is bifurcating into two distinct business models: energy-infrastructure companies that lease megawatts to hyperscalers, and pure-play miners running increasingly thin margins on block rewards alone. The economics of each diverge sharply.
Publicly listed miners—including MARA Holdings, CleanSpark, Riot Platforms, Core Scientific, Bitdeer, and Cango—collectively sold over 32,000 BTC in Q1 2026, according to data compiled by CoinDesk and BeInCrypto. The figure surpasses total BTC sold by public miners across all of 2025 and sets a quarterly record.
The largest individual seller was MARA Holdings, which offloaded 15,133 BTC for approximately $1.1 billion in March alone. Riot Platforms sold 3,778 BTC, generating $289.5 million. Core Scientific liquidated approximately 1,900 BTC worth $175 million in January—part of a stated plan to sell nearly all of its 2,500 BTC holdings to fund AI infrastructure buildout.
The primary driver was margin compression. With Bitcoin averaging below $80,000 for much of Q1 and hashprice at multi-year lows, miners lacked the runway to hold. The sell-off was amplified by macroeconomic headwinds: U.S. tariff escalation and rising interest rate expectations pushed risk assets lower across the board.
Hashprice—the daily revenue earned per petahash of deployed mining capacity—peaked at approximately $63/PH/s/day in July 2025. It declined to $35–37 by November 2025, briefly recovered to $38–40 in late December, and then collapsed to $28–30 by early March 2026, per CoinShares' Q1 2026 mining report.
At these levels, the economics are stark:
| Metric | Value | |---|---| | Weighted-average cash cost per BTC (Q4 2025, public miners) | ~$80,000 | | BTC spot price (May 1, 2026) | $78,178 | | Electricity-only breakeven (mid-tier operators) | ~$74,000 | | Efficient fleet breakeven (S21 XP, sub-$0.05/kWh power) | ~$55,000 | | Legacy fleet shutdown threshold | ~$75,000 | | Global fleet unprofitable at current levels | 15–20% |
CoinShares estimates that 15–20% of the global Bitcoin mining fleet is unprofitable at current hashprice levels, concentrated among operators running pre-S21-generation hardware or paying above $0.08/kWh for electricity. Industrial miners with sub-$0.05/kWh power retain approximately 3–4x the gross margin of retail miners paying the $0.16/kWh U.S. residential average.
The cost divergence between efficient and legacy operators continues to widen. Fleet efficiency for top-tier operators has improved to 16 J/TH (per CleanSpark's March 2026 data), while sub-15 J/TH machines now represent the minimum viable hardware class for positive margins at prevailing hash economics.
The most consequential structural shift in the mining sector is not about Bitcoin at all. Over $70 billion in cumulative AI and HPC hosting contracts have been announced across the public mining sector, according to S&P Global and Insights4VC research.
The logic is straightforward: miners control scarce assets that hyperscalers need—permitted sites with industrial power capacity, grid interconnection agreements, cooling infrastructure, and experience operating at scale in remote locations. Converting these assets from proof-of-work computation to GPU-based AI inference and training commands higher margins and multi-year revenue visibility.
By December 2026, publicly listed Bitcoin miners could derive up to 70% of their total revenue from AI and HPC operations, up from approximately 30% at the start of the year, according to S&P Global estimates. Three companies illustrate the scale of transformation:
IREN: HPC revenue projected to reach 71% of total revenue, up from 3% in 2024. Secured a $9.7 billion contract with Microsoft for 76,000 NVIDIA GB300 GPUs across 200 MW at its Childress, Texas campus. Q1 FY2026 revenue hit $240.3 million (up 355% YoY).
Core Scientific: Colocation revenue of $31.3 million in Q4 2025, up from $8.5 million a year prior. Over halfway through its 590 MW, $10–12 billion CoreWeave partnership. Issued a $3.3 billion bond in April 2026 to fund further AI expansion.
TeraWulf: Secured $12.8 billion in contracted HPC revenue with Google-backed Fluidstack and Core42 across sites totaling over 1 GW. HPC now accounts for over 50% of quarterly revenue. Raised $900 million in April 2026 equity offering.
The divergence in investor sentiment is measurable. Mining stocks as a group outperformed Bitcoin by 70% year-to-date, but the gains are concentrated among AI-pivoting operators:
| Company | Ticker | YTD Return | Primary Revenue Driver | |---|---|---|---| | TeraWulf | WULF | +73.58% | AI/HPC contracts | | Riot Platforms | RIOT | +47.04% | Hybrid (mining + data center) | | CleanSpark | CLSK | +25.88% | Pure-play mining | | MARA Holdings | MARA | -28.00% | Pure-play mining |
The total market capitalization of the public Bitcoin mining sector stands at $77.6 billion—1.8% of Bitcoin's market capitalization. The sector's valuation increasingly reflects contracted AI revenue multiples rather than Bitcoin mining economics. Investors are pricing in data center cash flows at enterprise infrastructure multiples, not mining hashrate.
Bitcoin's network hashrate crossed 1 Zetahash/s (1,000 EH/s) in January 2026, with the 7-day moving average reaching 1.05–1.13 ZH/s. In February, mining difficulty surged 15% to 144.4T—the largest single adjustment since 2021. Hashrate is projected to reach 1.8 ZH/s by year-end.
The paradox: network security is at an all-time high, but fee revenue is not keeping pace. Bitcoin transaction fees totaled approximately $115 million annualized as of the foundational economic value report, while block subsidy issuance was $18.1 billion annually. The halving in April 2024 cut block rewards from 6.25 BTC to 3.125 BTC, halving the primary revenue source while difficulty continued to climb.
This dynamic produces a structural squeeze: more hashrate competing for fewer rewards per block. The only relief valves are higher BTC prices, higher transaction fees, or operator capitulation. Q1 2026 saw the third option play out at scale.
Texas SB 6: Signed June 2025, the law requires facilities consuming 75+ MW to maintain curtailment readiness, allowing ERCOT to order shutdowns during grid emergencies. This directly affects mining operations at scale—and the AI data centers many miners are building. The provision was shaped by Winter Storm Uri, during which the state reportedly paid miners $175,000/hour to curtail.
Energy competition: AI data centers and Bitcoin miners now compete for the same scarce resource—permitted, grid-connected power at scale. As hyperscalers bid up power purchase agreements, pure-play miners face rising site acquisition costs even as their revenue per megawatt declines.
Hardware cycle: The efficiency floor keeps dropping. Sub-15 J/TH machines are now baseline. Operators running anything above 20 J/TH face negative unit economics at current prices, accelerating hardware obsolescence cycles and capex requirements.
Public miners sold a record 32,000 BTC in Q1 2026, exceeding all of 2025's liquidations, driven by hashprice at $28–30/PH/s/day—a level last seen in 2018.
The weighted-average cash cost per BTC for public miners reached ~$80,000, approximately equal to Bitcoin's spot price, leaving zero margin for operators without sub-$0.05/kWh power or latest-generation hardware.
Over $70 billion in AI/HPC contracts have been signed by former pure-play miners. Hut 8 ($7B Anthropic deal), Core Scientific ($10B+ CoreWeave), TeraWulf ($12.8B multiple counterparties), and IREN ($9.7B Microsoft) lead the transition.
Stock performance bifurcated sharply: AI-pivoting miners (TeraWulf +73.58%, Riot +47.04%) outperformed pure-play operators (MARA -28%). The market is repricing the sector as energy infrastructure, not Bitcoin proxies.
Bitcoin's hashrate hit 1+ ZH/s while fee revenue remains structurally low at ~$115M annualized. The network's security model remains subsidy-dependent, consistent with the broader finding that 85–90% of blockchain economic flows are subsidy-driven.
15–20% of the global mining fleet is currently unprofitable, per CoinShares. Further difficulty increases and flat BTC prices would widen this band.
The Bitcoin mining sector in Q1 2026 demonstrated the economic pressures that the post-halving environment imposes on operators dependent on block rewards. The record liquidation of 32,000 BTC represents forced selling by entities whose cost structures exceeded their revenue at prevailing prices—a textbook margin call across an entire industry segment.
The more consequential development is the structural transformation underway. Former mining companies are converting their core assets—power capacity, cooling infrastructure, permitted sites—into AI data center operations. The contracted revenue base of $70+ billion dwarfs the entire annualized Bitcoin mining revenue pool. These are not speculative pivots; they are backed by 10–15 year leases with Google, Microsoft, AMD, and Anthropic as counterparties or guarantors.
For the mining operations that remain, the economics are clear: sub-$0.05/kWh electricity and sub-15 J/TH hardware efficiency are the minimum viable thresholds. Operators above these lines face a choice between continued operation at a loss, liquidation of BTC reserves, or exit from mining entirely. The sector's subsidy dependence—$18.1 billion in annual block issuance supporting $115 million in fee revenue—remains the structural vulnerability that the halving cycle will continue to compress.
The market has already voted with capital flows: energy infrastructure with contracted AI revenue commands premium multiples, while pure-play mining carries the risk profile of a commodity producer without pricing power. Whether the remaining pure-play miners can survive until the next price cycle—or whether they too convert their megawatts to serve a different customer—will define the sector's composition by year-end.