The Squeeze Is Real
Two years after the April 2024 halving cut the block subsidy from 6.25 to 3.125 BTC, the economics of Bitcoin mining have reached a breaking point. CoinShares' Q1 2026 Bitcoin Mining Report lays out the numbers — and they are sobering.
Hash price, the metric that determines how much revenue a miner earns per unit of computational power, peaked at roughly $63 per petahash per second per day (PH/s/day) in July 2025. By early March 2026, it had collapsed to $28/PH/s/day — a five-year low. The weighted average cash cost to mine a single bitcoin among publicly traded miners rose to approximately $79,995 in Q4 2025, meaning many operators are now spending more to mine BTC than the coin is worth on the open market.
The result: roughly 15–20% of legacy mining machines worldwide are now cash-flow negative.
Miners Are Selling Their Bitcoin
When revenue can't cover operating costs, miners sell reserves. Publicly listed miners have collectively liquidated over 15,000 BTC from their treasuries since peak levels. Core Scientific sold approximately 1,900 BTC ($175 million) in January alone and signaled plans to sell substantially all remaining holdings during Q1. Bitdeer zeroed out its treasury in February. Riot Platforms sold 1,818 BTC ($162 million) in December 2025.
This is forced selling, not strategic rebalancing. When the entities responsible for producing new bitcoin are net sellers of existing supply, it tells you the margin environment is genuinely distressed.
Three Consecutive Negative Difficulty Adjustments
The on-chain data confirms the pain. Earlier in 2026, Bitcoin's mining difficulty posted three consecutive downward adjustments — the first such streak since July 2022. Network hashrate briefly dipped below the symbolic 1 zettahash per second mark on May 1, before recovering modestly.
Block times slowed, signaling that less-efficient miners were shutting down rigs. The May 15 adjustment brought difficulty back up 3.12% to 136.61 trillion, but the damage to the weakest operators was already done.
The AI Escape Hatch
Faced with sub-breakeven hash prices, the mining industry is doing what any capital-intensive business does when its core product stops paying: it's diversifying. And the diversification target is artificial intelligence.
CoinShares estimates that listed miners could derive as much as 70% of their revenue from AI and high-performance computing by the end of 2026, up from roughly 30% today. The collective value of GPU co-location and cloud service deals signed with hyperscalers since 2025 exceeds $70 billion.
TeraWulf exemplifies the trend. On May 26, the company announced a new 1-gigawatt data center campus in Kentucky aimed squarely at AI workloads — not Bitcoin mining. Its stock jumped 13% on the news. The facility, dubbed "Muskie Data Campus," will deliver an initial 500 megawatts starting in late 2028, with another 500 megawatts targeted by 2030.
The underlying logic is straightforward: these companies already own the real estate, the power contracts, and the cooling infrastructure. Plugging in GPUs instead of ASICs can generate higher and more stable revenue per megawatt.
What This Means for Bitcoin's Supply Dynamics
For long-term holders, the mining profitability crisis matters for two reasons.
First, forced miner selling adds supply pressure. When miners liquidate thousands of BTC to cover electricity bills and debt service, that bitcoin hits the open market. At current prices around $76,000, this is meaningful downward pressure — though it's also a finite process. Once weak miners exit, the selling stops.
Second, the AI pivot raises questions about network security incentives. If the most profitable use of mining infrastructure is running AI models rather than hashing blocks, capital allocation will follow. CoinShares forecasts the network will reach 1.8 ZH/s by year-end 2026 and 2 ZH/s by March 2027 — but that projection depends on BTC price recovering enough to make mining competitive with AI hosting.
The geographic concentration adds another layer. The top three countries — the United States, China, and Russia — now control approximately 68% of global hashrate, with the US gaining about 2 percentage points of market share quarter over quarter.
The Historical Pattern
Bitcoin mining has been through profitability crises before. The pattern is consistent: weak operators capitulate, difficulty adjusts downward, surviving miners become more profitable, and the network stabilizes at a new equilibrium. The 2022 cycle played out exactly this way after the Terra-Luna collapse and FTX bankruptcy.
What's different this time is the AI alternative. Previous cycles offered miners no viable exit — they either survived or shut down. Now they can repurpose infrastructure, which could accelerate the shakeout while simultaneously reducing the total hashrate available to the Bitcoin network.
Bitcoin Gate Take
This report confirms what on-chain data has been whispering for months: the post-halving margin compression is real and not yet fully resolved. The good news is that miner capitulation is historically a late-cycle bottoming signal — forced selling eventually exhausts itself, difficulty resets, and surviving operators enjoy wider margins. The bad news is that the AI pivot introduces a structural competitor for mining infrastructure that didn't exist in prior cycles. Watch the hashrate through Q3; if it stalls while price stays flat, the network's security budget conversation gets louder.
Bitcoin's retirement planning tools at bitcoingate.net factor in multiple growth models. Understanding supply dynamics — including miner behavior — helps you build more realistic long-term projections.