A race to the cheapest energy
Bitcoin mining is close to textbook perfect competition. The total revenue available to all miners is set by the Bitcoin price and the block reward — today almost entirely the fixed protocol subsidy, with transaction fees a low-single-digit share — and that pool is divided among everyone competing for it. Add a new machine and you take a slice of a fixed pie; the network's difficulty simply adjusts. The structural consequence is that margins compress over time, and the operators who endure are those with the lowest energy cost, often the ones monetizing power that would otherwise be wasted.
Selling Bitcoin is a symptom, not a surrender
That framework reframes the 32,000 BTC that listed miners sold in Q1 2026. By early 2026 the cost to mine a coin had risen near $90,000 against a Bitcoin price in the mid-$60,000s — mining margins were negative. In that state, a treasury of previously mined Bitcoin is working capital, and spending it to fund a transition toward cheaper or contracted revenue is a capital-allocation decision, not an ideological reversal. The hashrate's first Q1 decline since 2020 is the same story told through machines instead of coins.
Where CleanSpark fits
CleanSpark has run explicitly on this logic, balancing how much mined Bitcoin it retains against operating needs and converting low-cost power into longer-dated contracts. Under the last-survivor model, that is exactly the behavior the framework rewards: protect the cheapest, most flexible power, and treat the treasury as a resource rather than a monument. It is consistent with Signal21's constructive standing view — the caveat being that the same compression that pressures rivals never fully relents, so execution on power cost is the whole game. This is market analysis, not a recommendation to buy or sell any security.