Bitcoin Mining Has Entered a Much Harder Phase
Bitcoin mining is no longer moving through a normal cyclical slowdown. The sector has entered what looks like its harshest post-halving squeeze so far, with weaker economics, tighter margins, and a growing split between companies that still behave like pure miners and those trying to become AI and infrastructure businesses. A new CoinShares report frames Q4 2025 as the most difficult quarter for miners since the April 2024 halving, after a sharp Bitcoin drawdown collided with near-record hashrate and crushed profitability.
That matters because mining is no longer just a Bitcoin production story. It is becoming a capital allocation story. The key question is no longer only who can mine most efficiently, but who can survive margin compression while deciding whether power, land, and data-center capacity are more valuable for Bitcoin or for AI compute. That second point is an inference based on the report’s mining-cost data and its analysis of the sector’s AI migration.
Margins Are Under Real Pressure
The pressure is severe. CoinShares estimates that the weighted average cash cost to produce one Bitcoin among publicly listed miners rose to about $79,995 in Q4 2025. At the same time, hashprice sank to extreme lows, and the report notes that it fell even further than CoinShares had previously expected before stabilizing around a deeply compressed range.
This is the kind of setup that changes behavior fast. When production costs rise while hashprice stays weak, miners lose flexibility. Treasury selling becomes more likely, fleet quality matters more, and weaker operators stop looking like leveraged BTC plays and start looking like structurally stressed businesses. That interpretation is based on the economics described in the report rather than a direct sentence from CoinShares.
A Meaningful Slice of the Network Is Likely Underwater
The report’s most revealing point may be how much of the fleet is now vulnerable. At roughly $30 hashprice per PH/s/day, any miner operating below an S19 XP at power prices of 6 cents per kWh or more is losing money, according to CoinShares. The firm estimates that this applies to about 15% to 20% of the global mining fleet.
That helps explain why the network recently saw three consecutive negative difficulty adjustments, the first such streak since July 2022. CoinShares still expects the network to keep expanding over time and now projects hashrate reaching 1.8 zetahash by the end of 2026, but the path there looks much less smooth than it did a few months ago.
The Industry Is Starting to Split in Two
This is where the story gets bigger than miner pain. Public miners are no longer one coherent group. Some are still straightforward Bitcoin businesses. Others are becoming power-and-infrastructure companies with mining attached.
That shift is already visible in expected revenue mix. CoinShares estimates that listed miners could derive as much as 70% of revenue from AI by the end of this year, up from roughly 30% now. That is not a side experiment anymore. It suggests that, for a growing part of the sector, Bitcoin mining is becoming only one use of infrastructure rather than the entire business model.
The Market Is Rewarding the AI Pivot
Equity markets are already treating that split as real. CoinShares says miners with secured HPC contracts are trading at about 12.3x EV to next-twelve-month sales, compared with 5.9x for pure-play miners. That valuation gap is one of the clearest signs that investors are paying for future AI and compute revenue, not just exposure to Bitcoin block rewards.
That does not automatically mean the AI-pivot names are safer. It means the market believes their upside is broader. Bitcoin miners that can reposition themselves as data-center or compute providers are being valued more like infrastructure assets than simple cyclical commodity producers. That is an inference from the valuation spread and the report’s discussion of revenue diversification.
The Upside Comes With a New Risk Profile
There is a catch, though. A mining company shifting into AI is not just becoming more diversified. It is often becoming more leveraged, more capital-intensive, and more execution-dependent.
That is the part many market participants miss. The old mining model was already volatile, but it was relatively easy to understand: hardware efficiency, energy prices, Bitcoin price, and treasury management. The hybrid AI-mining model adds another layer entirely: contract execution, financing pressure, customer concentration, data-center buildout, and much larger capex requirements. That risk-profile shift is an inference from the report’s analysis of AI migration, sector valuation, and rising infrastructure intensity.
What This Means for Bitcoin
For Bitcoin itself, the message is not necessarily bearish. The network can keep growing even if mining becomes more selective. In fact, a harder environment may leave the industry increasingly concentrated in the hands of operators with the cheapest power, the strongest balance sheets, or the most flexible infrastructure. CoinShares still expects long-term hashrate growth despite the current squeeze.
But it does mean mining equities are becoming less pure as Bitcoin proxies. Buying miners today may mean buying some mix of BTC exposure, debt risk, AI optionality, infrastructure execution, and balance-sheet stress. That is a very different proposition from the older market habit of treating miners as simple leveraged bets on Bitcoin. This is analytical interpretation based on the report’s cost, valuation, and AI-revenue data.
BTCUSA Insight
The most important takeaway from the new mining data is not just that margins are tight. It is that the post-halving squeeze is accelerating a much deeper identity shift across the sector. Bitcoin mining is no longer evolving as one business. It is splitting into two: low-cost BTC production on one side, and AI-linked infrastructure on the other.
That is why this report matters. The next winners in mining may not be the companies that produce the most Bitcoin, but the ones that control the most adaptable infrastructure when BTC economics and AI economics pull in different directions. That final sentence is an inference grounded in the report’s core findings.
