It has been two years since the April 2024 halving cut Bitcoin's block subsidy from 6.25 BTC to 3.125 BTC. The post-halving period has played out differently from what the standard miner-economics models predicted. The differences are interesting and tell us something durable about the structure of Bitcoin mining as a business.

What the Models Said

The conventional pre-halving analysis went something like this. A halving cuts block subsidy revenue in half overnight. Miners with all-in costs at or above the new break-even point are forced to shut off rigs. Hashrate falls, difficulty adjusts down, and surviving miners benefit from the resulting concentration. The cycle repeats with each halving until block subsidy becomes economically irrelevant and transaction fees dominate miner revenue.

That story has been told before each halving. After 2020, it broadly held. After 2024, it did not.

What Actually Happened

Hashrate did not fall meaningfully after the 2024 halving. It was flat for roughly six weeks, then resumed an upward trajectory that has continued essentially uninterrupted for two years. Network hashrate is now approximately 70% higher than at the halving moment.

Several factors explain why the conventional shutdown story didn't materialize.

Energy cost compression. The marginal cost of mining-grade electricity has compressed considerably. Miners with access to stranded gas, behind-the-meter renewable contracts, and dedicated nuclear power agreements have all-in costs well below what was assumed in pre-halving models. The price floor below which mining becomes uneconomic moved meaningfully lower.

Fee revenue. Transaction fees have grown to a more substantial share of miner revenue than expected. Inscription-driven activity in 2024-2025 generated periodic fee surges, but the more durable shift is the slow normalization of fees as a meaningful proportion of total miner income. In Q1 2026, fees averaged 12% of total block reward — still secondary to subsidy but materially higher than the 1-3% average of the 2020-2023 period.

Bitcoin price appreciation. The post-halving period coincided with Bitcoin appreciating from roughly $65,000 to its current level. Mining revenue in dollar terms has grown despite the subsidy cut, which is the simplest explanation for why hashrate kept climbing.

The Structural Shifts

What's structurally different in 2026 than in 2024 isn't primarily about cost curves or fee dynamics. It's about who is mining and why.

The publicly traded U.S. miners — Marathon, Riot, CleanSpark, Cipher, and a smaller cohort — now operate as quasi-Bitcoin treasury vehicles with mining revenue as their operational engine. Their balance sheets carry meaningful Bitcoin holdings; their stock prices correlate more tightly with Bitcoin spot than with their operational mining margins. They are, increasingly, leveraged Bitcoin proxies that happen to operate mining infrastructure.

The international fleet is more diverse than at any point in Bitcoin's history. Russia, Argentina, Paraguay, the UAE, Ethiopia, Bhutan, and a handful of other jurisdictions have grown meaningful hashrate footprints over the past two years. The geographic concentration of Bitcoin mining — which had moved heavily toward the U.S. after China's 2021 ban — has partially reversed.

The ASIC supply chain has consolidated further. Bitmain remains the dominant manufacturer, with MicroBT a meaningful second. Newer entrants have struggled to compete on power efficiency, and the manufacturing capacity is now structurally tilted toward the existing duopoly.

What's Different About 2028

The 2028 halving will cut subsidy from 3.125 BTC to 1.5625 BTC. Two structural questions for that event are worth thinking about now.

First: can fees grow to compensate? At current trajectory, fees would need to roughly double in dollar terms to keep total miner revenue flat post-2028. That's a substantial growth requirement for a metric that has been historically volatile. The candidates for sustained fee growth are: continued L2 settlement traffic on Bitcoin mainnet, ordinals-style protocol activity, and increased cross-chain bridging activity that relies on Bitcoin block confirmations. Each is plausible; none is certain.

Second: how concentrated does mining become? If hashrate growth continues to be driven by capital-rich publicly traded entities and well-funded sovereign-adjacent operators, the long-tail of small-scale mining will continue to compress. That has implications for the network's decentralization story, even if no individual operator approaches a controlling share.

The Read

The post-2024 halving period was not what the models predicted because the models underweighted Bitcoin price appreciation, energy cost compression, and the structural maturation of mining as a capital-allocation business rather than a pure operating business. Two years later, the network is materially larger and more institutionalized, and the next halving will arrive in a context that looks fundamentally different from any prior halving.

What hasn't changed: every four years, the subsidy gets cut in half. Whether the network keeps absorbing those cuts gracefully is the open question. Two halvings of grace are not three.