A Bitcoin halving cuts the block subsidy — the newly minted coins a miner is allowed to pay itself — in half. It does not touch transaction fees, and it does not directly touch difficulty, hashrate or price; those move on their own schedules, driven by market conditions rather than by the halving itself.
What actually gets cut
Every block a miner produces contains a coinbase transaction, the first transaction in the block, where the miner pays itself. That payment has two components, according to Cube Exchange: the block subsidy, which is newly created bitcoin, and transaction fees, which users attach to their transactions and which the miner simply collects. Halving affects only the subsidy component. Fees are never halved by the protocol.
At block height 840,000, this rule cut the subsidy from 6.25 BTC to 3.125 BTC, according to Blockstream, Cube Exchange, Kraken and Spark Money, each reporting the same subsidy figures independently. The calendar date is not fully agreed across these sources: Blockstream’s own halving table gives 19 April 2024, while Kraken and Spark Money give 20 April 2024. The underlying protocol event is a block height, not a calendar date, which is one reason outlets writing about it after the fact can land a day apart. Cube Exchange walks through the arithmetic: every node divides the block height by 210,000 to count how many halvings have occurred — at height 840,000 that comes to four — then starts from the original 50 BTC subsidy and halves it four times, arriving at 3.125 BTC. Any coinbase transaction claiming more than that amount, plus fees, is rejected by the network. A miner can claim less than the maximum, but never more.
Why revenue does not simply halve
Total block reward is subsidy plus fees, so total miner revenue depends on how large the fee component is at any given time. Cube Exchange notes that block 840,000 itself paid out well above the new 3.125 BTC subsidy because it happened to include unusually high transaction fees. That is the mechanism working exactly as designed: the subsidy is fixed by rule, the fee share floats with network demand for block space.
After the 2024 halving, fees became a smaller share of total block reward, not a larger one. Blockspace, in a one-year-later assessment cited by TheCoinrise’s reference desk, reported that transaction fees fell from 7.12% of block reward before the halving to 4.02% after it — a single-sourced figure with no stated methodology in the material reviewed for this page. Combined with a subsidy that was already halved, that meant less cushion for miners from fee income at the exact moment the subsidy shrank.
Difficulty and hashrate: a separate mechanism
Bitcoin targets one block roughly every 10 minutes. Every 2,016 blocks — about two weeks — the protocol compares actual block production time to the expected schedule and adjusts mining difficulty accordingly, according to both Fidelity Digital Assets and Spark Money. This retarget has nothing to do with the halving interval of 210,000 blocks; it runs on its own two-week cycle whether or not a halving has just occurred.
This is why hashrate does not automatically drop when a halving cuts revenue. Fidelity Digital Assets, in a March 2024 report, found that network hashrate climbed 8000% since the 2016 halving and 394% since the 2020 halving, as measured at the end of 2023 — a period that included a 40% hashrate decline in 2021 following China’s mining ban, and a bitcoin price crash from a peak of $67,000 to $15,000. Hashrate kept climbing through both events. Miners who can no longer operate profitably shut down and difficulty falls to compensate; the network does not lose block production capacity just because per-block revenue fell.
A worked example of hash price
Fidelity Digital Assets illustrates how a miner converts hashrate into expected income using “hash price,” the going rate the network pays per unit of computing power. In its March 2024 example, a single machine producing 110 terahashes per second could expect to earn about $9.51 a day, calculated against a hash price of $0.086 per terahash at that time. This was a pre-halving snapshot; hash price itself is not fixed by protocol and changes with bitcoin’s price, transaction fee volume and total network hashrate.
What happened to hash price after April 2024
Blockspace’s post-halving assessment, summarized by TheCoinrise’s reference desk, reported that USD-denominated hash price fell from an average of about $66.35 per petahash per day between April 2023 and April 2025 to about $45 per petahash per day as of July 2025. In bitcoin terms, Blockspace put the decline at 0.001462 BTC per petahash per day down to 0.000517 BTC per petahash per day over the same window, which the outlet summarized as a 57% fall in USD hash price and an 89% fall in BTC-denominated hash price since April 2024. These are Blockspace’s own figures, cited by a single outlet in the evidence reviewed for this page, with no disclosed data source or calculation method.
Spark Money, in a March 2026 analysis, put the average cash cost to mine one bitcoin at approximately $74,600 as of late 2025, rising to roughly $137,800 once depreciation, financing and stock-based compensation are included. Spark Money set this against a bitcoin price fluctuating between $78,000 and $126,000 during the same period, and described margins for all but the most efficient miners as thin. The outlet also estimated that producing one bitcoin now requires about 854,400 kilowatt-hours of electricity, and that electricity makes up 75-85% of monthly mining costs. None of these figures are attributed to a named primary data source or stated calculation method within Spark Money’s article.
Newly issued bitcoin, which is the volume miners have to sell into the market each day, fell from roughly 900 BTC per day before the 2024 halving to roughly 450 BTC per day afterward, according to both Blockstream and Bitcoin Foundation. That is a direct, protocol-verifiable consequence of the subsidy change, unlike the hash price and cost figures above.
The common misreading
Readers often assume the halving cuts a miner’s total revenue in half. It cuts only the subsidy in half; total revenue is subsidy plus fees, and actual profitability also depends on bitcoin’s price and on network difficulty at the time, none of which the halving controls directly. A related error is assuming hashrate falls in response to a halving. As Fidelity Digital Assets documents, hashrate kept rising through the 2020 halving, a 40% mining-ban shock in 2021, and a bear market that took bitcoin from $67,000 to $15,000 — because difficulty adjusts to how fast blocks actually arrive, not to whether mining is profitable.
What this page does not tell you
This page cannot tell a reader today’s hash price or current mining margin. The figures cited here are dated snapshots: Fidelity’s worked example is from March 2024, Spark Money’s cost figures are described only as “as of late 2025,” and Blockspace’s hash price figures run through July 2025. Hash price moves continuously with bitcoin’s price, network difficulty and fee volume.
Spark Money’s cash-cost and all-in-cost figures are self-reported by that outlet with no cited primary source or stated methodology in the material reviewed here, so this page cannot verify how those costs were calculated or which miners were sampled to produce them. Blockspace’s fee-share and hash price decline figures come from a single outlet in this evidence set and are likewise presented without a disclosed calculation method.
This page also does not address what happens to Bitcoin’s security budget as the subsidy approaches zero, expected around the year 2140 according to Blockstream’s reading of Bitcoin Core’s source code — that is a longer-horizon question distinct from what changes at any single halving. Finally, this page describes the mechanism and what was reported after the 2024 halving; it makes no claim about what will happen to miner revenue, hash price or bitcoin’s price around the fifth halving, projected for block height 1,050,000 in roughly March or April 2028 according to Spark Money and Kraken.
Sources
Every fact above is attributed to one of these reports. Where they disagree, the article says so.
