Bitcoin mining is experiencing its most severe economic stress since the 2022 bear market. With the spot price hovering near $68,000 — roughly 20% below the industry-average all-in production cost of $87,000 — the majority of the global hashrate is operating underwater for the first time in four ...
"We're not a bitcoin company anymore." — Bitfarms CEO Ben Gagnon, announcing the company's full pivot to AI data centers, February 2026
Bitcoin mining is experiencing its most severe economic stress since the 2022 bear market. With the spot price hovering near $68,000 — roughly 20% below the industry-average all-in production cost of $87,000 — the majority of the global hashrate is operating underwater for the first time in four years. The result is a rapid Darwinian shakeout: miner outflows spiked to 28,605 BTC ($1.8 billion) in a single day on February 5, difficulty posted its largest drop since China's 2021 ban, and at least eight publicly traded miners have announced partial or complete pivots to AI infrastructure.
Yet the network itself is proving remarkably resilient. After plunging from 1.1 ZH/s to 826 EH/s during a severe U.S. winter storm, hashrate has recovered to 1 ZH/s — triggering a 15% upward difficulty adjustment on February 20, the largest since 2021. The paradox is clear: the network is getting harder to mine at the exact moment mining is least profitable. What is unfolding is not a temporary dip but a structural repricing of who can afford to secure Bitcoin and what the survivors will look like.
The economics of Bitcoin mining in February 2026 are brutally simple. According to Checkonchain data, the industry-average all-in cost to produce one bitcoin sits at approximately $87,000. The spot price: ~$68,000. That is a 22% deficit — the widest negative margin since the depths of the 2022 bear market.
But averages obscure the real story. The cost structure across the industry is wildly bifurcated:
| Miner Tier | Estimated Cost/BTC | Status at $68K | |---|---|---| | Tier 1 (MARA, CLSK) | $34,000–$43,000 | Profitable | | Tier 2 (Mid-scale, mixed energy) | $55,000–$70,000 | Marginal | | Tier 3 (Small-scale, retail power) | $75,000–$120,000+ | Underwater | | Industry Average | ~$87,000 | Underwater |
Hashprice — the estimated daily revenue per petahash of mining capacity — has collapsed to $23.9/PH/s, down roughly 35% from its October 2025 peak of ~$55/PH/s when Bitcoin hit its all-time high near $126,500. At current levels, the return-on-investment timeline for new mining hardware now exceeds 1,000 days — approaching three years for a machine with an expected operational lifespan of 18–24 months. The math has stopped working for most participants.
Electricity, which represents 60–80% of operating costs, is the decisive variable. At $0.06/kWh — a competitive but not exceptional rate — mining one bitcoin now requires approximately 854,400 kWh and costs roughly $51,264 in electricity alone. Add hosting, maintenance, depreciation, debt service, and corporate overhead, and the all-in figure balloons to $75,000–$87,000 for most publicly traded operators.
The economic stress was compounded by a severe winter storm that swept across the U.S. in late January 2026, forcing emergency power curtailments across the ERCOT grid in Texas — the epicenter of North American mining. The impact was immediate and historic:
On February 9, Bitcoin's mining difficulty posted an 11% downward adjustment to ~125 T — the largest single decline since China's mining crackdown in mid-2021. The drop confirmed what on-chain data had been signaling for weeks: miners were not just curtailing temporarily — many were shutting down permanently.
But here is where the story takes a counterintuitive turn. By mid-February, hashrate had snapped back to 1 ZH/s, triggering a massive 15% upward difficulty adjustment to 144.4 T on February 20. Efficient miners, sensing opportunity in the difficulty drop, had aggressively deployed new capacity to capture a temporarily larger share of block rewards. The result: difficulty is now higher than it was before the storm, creating an even more punishing environment for marginal operators.
The financial stress has manifested in on-chain data as a textbook capitulation event:
CryptoQuant's miner capitulation indicator, which tracks the ratio of miner revenue to the 365-day moving average, has entered the "capitulation zone" for the first time since mid-2024. Checkonchain described the movement as a "rapid capitulation event with high volume and large realized losses."
The pain is not equally distributed. Among publicly traded miners:
The distinguishing feature of this cycle's capitulation is the scale of corporate Bitcoin treasuries at risk. Publicly traded miners — MARA, Riot, CleanSpark, IREN — collectively hold massive BTC positions financed in part by debt. If forced liquidation of these treasuries begins in earnest, the supply-side impact on Bitcoin's price could create a reflexive downward spiral that the 2022 cycle's mostly private miners never posed.
Confronted with margins that no longer justify their core business, the mining industry is undergoing its most dramatic strategic reorientation in history. At least eight publicly traded Bitcoin miners have announced partial or complete pivots to AI and high-performance computing (HPC) infrastructure:
| Company | Strategy | Status | |---|---|---| | Bitfarms | Full exit from BTC mining; rebranding as "Keel Infrastructure" | Winding down mining over 2 years | | Core Scientific | Converting facilities to AI hosting for CoreWeave | Operational | | CleanSpark | Acquired 447 acres in Texas for AI data center (600 MW) | Development phase | | IREN | Dual BTC/AI operations | Stock up 357% YTD | | TeraWulf | Pivoting Lake Mariner facility to HPC | Transitioning | | Hut 8 | GPU-as-a-service for AI workloads | Operational | | Bit Digital | AI/HPC revenue now exceeding mining revenue | Operational | | Cipher Mining | Data center conversion program | Planning |
The financial logic is compelling. AI hyperscalers — Microsoft, Amazon, Google, Meta — are collectively spending over $500 billion on data center infrastructure. Bitcoin miners possess the exact assets these companies need: large-scale power interconnections, cooling infrastructure, and land in energy-rich jurisdictions. A megawatt of capacity hosting AI workloads can generate 3–5x the revenue of a megawatt deployed for Bitcoin mining, with far more predictable cash flows.
The stock market has noticed. IREN, which committed earliest and most aggressively to AI, has delivered a 357% return year-to-date. TeraWulf is up 154%. Meanwhile, MARA — the most committed Bitcoin-only miner among the majors — is the only company in negative territory, down 28% YTD.
The market is sending an unmistakable signal: investors value the infrastructure, not the mining.
The miners who will survive this shakeout share three common characteristics:
1. Sub-$45,000 production costs. Only Tier 1 operators with owned or contracted power at $0.03–$0.04/kWh can remain profitable at current prices. MARA and CleanSpark have demonstrated production costs in the $34,000–$43,000 range — roughly half the industry average. Their advantage comes from vertically integrated power generation, often from renewable sources (hydro, wind, flared gas) where marginal electricity costs approach zero.
2. Diversified revenue streams. The companies being rewarded by the market are those treating their power and land assets as a platform, not a single-purpose mining operation. Core Scientific's pivot to hosting CoreWeave's AI workloads transformed it from a distressed miner into an infrastructure company valued on enterprise contracts rather than hashprice.
3. Balance sheet discipline. The 2025 bull run tempted many miners to leverage up, issuing convertible notes and taking on debt to acquire more machines and more Bitcoin for their treasuries. Those bets are now unwinding. The survivors are those who maintained conservative debt-to-equity ratios and avoided the siren call of maximum BTC accumulation at all-time-high prices.
The concentration of mining power among a shrinking number of well-capitalized operators raises legitimate questions about Bitcoin's security model. If mining continues to consolidate around a handful of companies that view it as a secondary business line behind AI hosting, the network's censorship resistance and decentralization — the properties that justify its existence — may gradually erode.
More immediately, the subsidy economics remain stark. Bitcoin's annualized fee revenue sits at roughly $115 million — a rounding error compared to the $18.1 billion in annual block reward issuance that actually pays miners. The network's security budget is almost entirely dependent on token inflation, not organic demand for block space. Each halving makes this tension more acute, and the April 2024 halving's effects are now fully visible in the wreckage of Q1 2026.
The mining industry is consuming an estimated 175 TWh of electricity annually — comparable to Poland's entire national consumption — to secure $115 million in fee revenue. Whether that energy expenditure is justified by Bitcoin's broader value proposition as a monetary network is a question the market is currently answering with mining stock prices.
The Bitcoin mining industry of February 2026 is undergoing a forced evolution. The combination of post-halving revenue compression, a 44% price decline from all-time highs, extreme weather disruptions, and competition from AI for both electricity and investor capital has created a selection pressure that is rapidly reshaping the sector.
What emerges from this crucible will be a smaller, more concentrated, and more diversified mining industry — one where Bitcoin mining is increasingly a feature of a broader infrastructure business rather than a standalone enterprise. The companies that survive will be those that recognized earliest that their real asset was never hashrate. It was power.
For Bitcoin's network, the implications are more ambiguous. A more professionalized, infrastructure-oriented mining sector may be more efficient, but it is also more correlated with traditional financial markets, more responsive to regulatory pressure, and less ideologically committed to the decentralization principles that mining was designed to protect. The miners are evolving. Whether Bitcoin's security model can evolve with them remains the trillion-dollar question.