Public Bitcoin Miners Went Unprofitable This Year
The public bitcoin mining sector posted its first-ever collective loss in the second quarter of 2026, according to a new CoinShares report, with the average cost of mining a single bitcoin running roughly 29% above where the coin actually traded by quarter’s end. It’s a genuinely unprecedented moment for an industry that has weathered plenty of volatility before, but never quite this specific kind of structural squeeze.
The Numbers
CoinShares’ research puts the average all-in cost of mining one BTC across public mining companies at approximately $75,500 during the second quarter. Bitcoin itself closed the quarter at $58,400, a gap of nearly $17,000, or about 29% below what it actually cost these companies to produce each coin. That’s the first time, in the history of the public mining sector, that the group has collectively fallen below the breakeven line on a quarterly basis.
The response has already started showing up in hashrate plans. At least 35 EH/s of computing capacity is now scheduled to be withdrawn from public miners specifically, roughly 4.7% of the network’s current total hashrate of around 750 EH/s. That’s a meaningful chunk of capacity, and it points to real operational decisions being made in response to the economics, not just a temporary dip in reported margins.
Why This Happened
The core dynamic is straightforward: mining costs are largely fixed in the short term, energy contracts, hardware, facility overhead, while bitcoin’s price is not. When price falls below the cost floor for a sustained period, miners without meaningful reserves or diversified revenue face a genuine profitability crisis, not just a bad quarter on paper.
CoinShares’ analysts frame bitcoin’s own price as the dominant factor shaping mining economics going forward, more than any single operational or hardware variable. A related dynamic they point to: hashrate itself is increasingly changing hands rather than growing net-new. Struggling or unprofitable operations aren’t simply shutting down capacity outright in every case; a meaningful share of that capacity is being sold, restructured, or acquired by better-capitalized players instead, which shows up in the data as capacity moving between companies rather than disappearing from the network entirely.
Geographically, the concentration at the top of the mining map remains pronounced. The United States, China, and Russia together control roughly 68% of total network hashpower, with the U.S. specifically extending its lead over the period. Beyond the top three, the report highlights a set of smaller, emerging markets building real mining capacity, including Paraguay, where HIVE is running a 300-megawatt project, Ethiopia, where Bitdeer operates a 40-megawatt facility, and Oman, all newer entrants capitalizing on relatively low energy costs at a moment when energy economics matter more than ever for anyone trying to mine profitably.
Who’s Positioned to Survive, and Who Isn’t
Not every public miner is facing this squeeze equally, and the report draws a fairly clear line between two groups.
Miners that have already pivoted meaningful infrastructure and capacity toward AI compute workloads are expected to stay concentrated in that segment rather than returning to pure bitcoin mining, a trend that’s been building for a couple of years as AI compute demand has offered better margins than bitcoin mining alone during periods exactly like this one. Hut 8’s modular infrastructure model, introduced back in March, which lets facilities switch compute capacity between AI workloads and bitcoin mining depending on which is more economical at a given time, is a specific example of the kind of flexibility that’s increasingly separating resilient operators from exposed ones.
CoinShares expects new investment in expanding pure bitcoin mining capacity to come mainly from the more flexible operators still active in the space, naming Riot, MARA, HIVE, and Bitdeer specifically as companies that have retained the ability to operate across both bitcoin mining and other compute workloads rather than being locked into bitcoin mining exclusively. That flexibility looks increasingly like the deciding factor in this environment: companies that can shift capacity toward whichever workload is actually profitable in a given quarter have a structural advantage over those that can’t.
What This Means Going Forward
A sector-wide negative margin quarter is a real signal, not just a headline number. It typically precedes consolidation, hashrate shifting from weaker to stronger operators, and, in some cases, outright capacity withdrawal from the network entirely rather than a temporary pause. The 35 EH/s already earmarked for removal from public miners specifically suggests that process is already underway rather than merely a risk on the horizon.
For bitcoin’s broader security model, a hashrate contraction of this scale, even a mid-single-digit percentage of total network hashpower, is worth watching, though it’s not automatically alarming on its own; mining difficulty adjusts to reflect actual network hashrate, and periods of miner consolidation have occurred before without threatening the network’s underlying security. What this quarter’s numbers do make clear is that the economics of public bitcoin mining specifically have shifted meaningfully, and the operators best positioned to weather that shift are the ones that built in flexibility, whether toward AI compute or simply stronger balance sheets, well before the squeeze actually arrived.