Skip to content
Menu
Finabase Learn

Crypto · How blockchains work

Bitcoin mining and the halving: rewards, difficulty and price

How miners earn bitcoin, how difficulty keeps blocks ten minutes apart, the halving schedule computed block by block, and what price did around three halvings.

On this page
  1. What miners do
  2. The halving schedule, computed
  3. Difficulty: why more miners do not mean more coins
  4. The economics of a miner
  5. Halvings and the price: what the data shows
  6. Pitfalls
  7. How Finabase does it
  8. Frequently asked questions
  9. Summary
  10. Check your understanding

Bitcoin mining is the work of adding blocks to bitcoin's ledger, and the halving is the rule that cuts the reward for that work in half every 210,000 blocks, about every four years. Together they do two jobs. Mining secures the record of who owns what. The reward schedule decides how many new coins come into existence and when.

The schedule is the reason people say there will only ever be 21 million bitcoin, and the halvings are the events around which a great deal of market folklore has grown. This lesson computes the schedule from the code's own constants, works through how the network keeps blocks ten minutes apart whatever miners do, shows what a miner's income looks like, and then checks what the price did around the three halvings in our data.

What is bitcoin? How it works, its supply cap and its price history introduces the coin itself. This page goes into the machinery.

What miners do

Every ten minutes or so, one miner somewhere wins the right to add the next block. To win, the miner must find a block header whose hash is below the network's target, which can only be done by trying an astronomical number of candidates. Proof of work vs proof of stake: how they differ runs a small version of that search.

The winning block contains one special transaction, placed first, with no sender. It is called the coinbase transaction, and it pays the miner two amounts:

  • the block subsidy, which creates new bitcoin; and
  • the transaction fees of every transaction in the block.

This is the only way bitcoin is ever created. There was no sale and no allocation to founders: every coin in existence was first paid to a miner. Coins from a coinbase transaction cannot be spent until 100 more blocks have been added, a rule that protects against rewards from blocks that end up discarded.

The whitepaper sets out the logic. The subsidy gives miners a reason to support the network while it is young and adds coins to circulation without a central issuer. Over time the subsidy shrinks and fees are meant to take over as the miners' income.

The halving schedule, computed

The rule in Bitcoin Core is short. The subsidy starts at 50 BTC and is cut in half, in whole satoshis (hundred-millionths of a bitcoin), once for every 210,000 blocks that have passed. Running that rule forward gives the whole monetary history of bitcoin.

Era Starts at block Began Subsidy Coins issued in the era Total issued by its end
1 0 January 3, 2009 50 BTC 10,500,000 10,500,000 (50.00%)
2 210,000 November 28, 2012 25 BTC 5,250,000 15,750,000 (75.00%)
3 420,000 July 9, 2016 12.5 BTC 2,625,000 18,375,000 (87.50%)
4 630,000 May 11, 2020 6.25 BTC 1,312,500 19,687,500 (93.75%)
5 840,000 April 20, 2024 (UTC) 3.125 BTC 656,250 20,343,750 (96.88%)
6 1,050,000 about April 2028 1.5625 BTC 328,125 20,671,875 (98.44%)
7 1,260,000 about 2032 0.78125 164,062.5 20,835,937.5 (99.22%)

Dates are the timestamps of the halving blocks in UTC. The 2024 halving fell on the evening of April 19 in US Eastern time.

Each era issues half as many coins as the one before, so the total approaches a limit without the code ever naming it. After 33 eras the subsidy rounds down to zero satoshis, around the year 2140, and the sum of everything issued is 20,999,999.9769 BTC, just under 21 million. Half of all bitcoin was issued in the first four years. The last 1% will take more than a century.

Bitcoin’s reward per block and total supply, by block height 0 0.00M 12.5 5.25M 25 10.50M 37.5 15.75M 50 21.00M Reward per block (BTC) Coins issued so far 0 210k 420k 630k 840k 1050k 1260k 1470k 1,680k November 2026: about block 976,000 2009 2012 2016 2020 2024 2028* 2032* 2036* Reward per block (left) Total coins issued (right) Block height on the horizontal axis, with the year each era began. * expected, at ten minutes a block.
Bitcoin’s block subsidy and cumulative supply by block height, computed from the rule in Bitcoin Core. Years marked with an asterisk are estimates at ten minutes a block.

Where the schedule stands

On November 18, 2026 the crypto list showed 20,113,584 BTC in circulation, 95.78% of the limit. That is 426,084 coins into the fifth era, which at 3.125 a block puts the chain about 136,350 blocks past the last halving, near block 976,350. A block explorer showed block 976,429 the same day.

That leaves roughly 73,600 blocks to the next halving at block 1,050,000. At ten minutes each that is about 511 days, which points to around April 2028. The date is an estimate: halvings are fixed by block count, not by the calendar, and blocks have historically arrived a little faster than every ten minutes because mining power has kept growing.

At 3.125 BTC a block and 144 blocks a day, 450 new bitcoin are created daily, about 164,000 a year. Against the existing supply that is an issuance rate of 0.82% a year, which will halve to about 0.4% in 2028.

Difficulty: why more miners do not mean more coins

If blocks are found by trial and error, then doubling the world's mining machines should find them twice as fast. For a short while it does. Then the network corrects.

Every 2,016 blocks, each node looks at how long the last 2,016 took. The intended time is two weeks: 2,016 blocks at ten minutes is exactly 14 days. The node then moves the target in proportion.

A worked example. Suppose new machines come online and the last 2,016 blocks took 12 days, an average of 8 minutes 34 seconds a block. The adjustment is 14 ÷ 12 = 1.167, so difficulty rises 16.7% and blocks return to ten minutes at the new, higher level of mining power. If instead miners switch off and the period takes 17.5 days, difficulty falls by 14 ÷ 17.5 = 0.80, a 20% cut, and the remaining miners find blocks at the usual pace again.

Three consequences follow.

  • Issuance cannot be sped up. No amount of investment in mining produces more than the scheduled coins for long. The extra power only raises the bar for everyone.
  • Mining is zero-sum among miners. The daily prize is fixed at 450 BTC plus fees. A miner's share of it equals their share of the network's total computing power, so one miner's new machines reduce everyone else's income.
  • The system self-heals. If mining becomes unprofitable and miners leave, difficulty falls until it is profitable for those who remain. The adjustment is slow, though: with half the miners gone, the next 2,016 blocks would take four weeks to arrive.

The economics of a miner

A miner's income is denominated in bitcoin and its costs in ordinary money: machines, buildings, staff and above all electricity. That mismatch is the whole business.

Consider an invented miner with 0.5% of the network's computing power. Its expected share of the daily subsidy is 0.5% of 450, or 2.25 BTC. At the $64,188 bitcoin price of November 18, 2026 that is about $144,400 a day before fees. If its electricity and other running costs come to $110,000 a day, it keeps $34,400.

Now change one input at a time:

Change Daily income Daily cost Result
Base case $144,400 $110,000 +$34,400
Bitcoin falls 30% $101,100 $110,000 −$8,900
Network power rises 25% (the miner's share falls) $115,500 $110,000 +$5,500
A halving cuts the subsidy to 1.5625 BTC $72,200 $110,000 −$37,800

The numbers are illustrative; real miners differ mainly in what they pay for power and how efficient their machines are. The pattern is general. A miner's margin is squeezed by a falling price, by rivals adding machines and, on a known date, by the halving. Miners with the highest costs switch off first, difficulty then falls, and the survivors' share rises. This is why shares of listed mining companies tend to move with bitcoin and by more than bitcoin in both directions. They are a leveraged bet on the price with operating risk on top, not a substitute for the coin.

Because any one miner might wait months between blocks, nearly all mining is done through pools, which combine members' computing power and split rewards in proportion to work contributed. Pools smooth income. They also mean a few pool operators assemble most blocks, a concentration worth knowing about (see How a blockchain reaches agreement: consensus explained on what a majority of mining power can and cannot do).

What happens when the subsidy runs out

The design assumes that fees will replace the subsidy as it shrinks. Whether they will be large enough to pay for the same level of security is an open question, and it is bitcoin's main long-term uncertainty. Fees depend on demand for scarce block space, which has been strong in bursts and weak in between. Crypto transaction fees and gas: how they are set explains how the fee market works.

Halvings and the price: what the data shows

A common claim is that halvings cause bull markets, because the flow of new coins for sale drops while demand continues. The claim is hard to test, because there have been only four halvings and our price history starts in late 2014, after the first.

Here is what month-end closing prices did around the other three, with the close of the halving month set to 100.

Halving Close that month 12 months before 12 months after 18 months after 24 months after 30 months after Highest month-end
July 2016 $625 46 460 1,636 1,246 554 2,266
May 2020 $9,461 91 395 603 336 181 648
April 2024 $60,637 48 155 181 126 128 191
Bitcoin before and after three halvings (halving month close = 100) 25 50 100 200 400 800 1600 3200 −12 −6 halving +6 +12 +18 +24 +30 July 2016 halving May 2020 halving April 2024 halving Months from the halving BTC-USD month-end closes, July 2015 to October 2026. Log scale. Three events are too few to establish a pattern.
BTC-USD month-end closes rebased to 100 at each halving month. Price only; log scale.

What can honestly be said:

  • The price was higher a year after each halving: by 360%, 295% and 55%.
  • Each gain was smaller than the last. A peak of 22.7 times the halving price, then 6.5 times, then 1.9 times.
  • Each was followed by a large fall. Thirty months after the 2016 halving the price was 76% below its month-end peak; after 2020, 72% below. Thirty months after April 2024, in October 2026, bitcoin was 33% below its month-end peak and 28% above the halving price.
  • The price had already doubled in the year before two of the three. A halving is the most predictable event in finance: its block number has been public since 2009. An efficient market should not be surprised by it.

What cannot be said is that halvings caused these moves. Three observations cannot separate a halving effect from everything else that was happening: interest rates near zero in 2020 and 2021, the launch of US spot bitcoin funds in January 2024, and the general tendency of a small, volatile asset to move in large cycles. The shrinking size of each move fits the arithmetic too. The 2016 halving removed 1,800 new coins a day from a supply of 15.75 million; the 2024 halving removed 450 a day from 19.7 million. Each halving cuts a smaller flow relative to what already exists and already trades.

The pattern is a description of three past episodes. It is not a forecast, and anyone who bought on the strength of it after the 2016 or 2020 halving and held through the following two years lived through a fall of more than 70%.

Pitfalls

  • Treating the halving date as fixed. It is a block height. The calendar date moves with the pace of mining and is only known to within days until shortly before.
  • Confusing supply cap with price floor. A fixed supply says nothing about demand. An asset can be scarce and fall 75%, as bitcoin did from December 2017 to January 2019.
  • Assuming mining is passive income. Mining bitcoin at a profit takes specialized machines and cheap power. An offer of effortless mining income deserves the same scrutiny as any other promised return.
  • Reading miner stocks as bitcoin. They carry debt, dilution, power contracts and management risk that the coin does not.
  • Forgetting that lost coins stay lost. The 21 million limit is an upper bound on issuance. Coins whose keys are lost can never move, so the usable supply is lower by an amount nobody can measure exactly.

How Finabase does it

Finabase shows bitcoin as BTC.X, with its circulating and maximum supply on the crypto list and a chart on the chart tab. It does not show block height, difficulty, mining power, fees or a halving countdown; those come from the network and are published by block explorers. Listed mining companies are ordinary stocks on Finabase with financial statements, and the comparison tab can plot one against bitcoin, for example MARA against BTC.X.

The schedule on this page was computed by the figure script from the constants in Bitcoin Core (50 BTC, 210,000 blocks, whole satoshis). The price table uses Finabase's monthly BTC-USD history, which begins in October 2014, so the November 2012 halving is not covered. The supply and price are those of November 18, 2026.

Frequently asked questions

When is the next bitcoin halving?

The next halving happens at block 1,050,000, when the subsidy falls from 3.125 to 1.5625 BTC. From the chain height on November 18, 2026 and ten minutes a block, that is about 511 days away, around April 2028. The exact date depends on how fast blocks arrive.

How many bitcoin are left to mine?

About 886,000 as of November 18, 2026: the limit of just under 21 million less the 20.11 million already issued. Because each era issues half as much as the last, these will be released over more than a century.

Does the halving make the price go up?

Nobody can show that it does. The price was higher a year after each of the 2016, 2020 and 2024 halvings, by progressively less, and fell heavily afterward each time. Three events with many other things going on are not enough to establish cause.

Can the 21 million limit be changed?

Only if the people running bitcoin nodes chose to adopt software with a different rule. Technically it is a few lines of code. In practice a change that diluted every holder would be rejected by nearly all of them, which is what makes the limit credible.

Is bitcoin mining profitable?

For industrial operators with cheap power and efficient machines, in some periods. Income is set by the bitcoin price and the miner's share of total computing power, both outside its control, while costs are fixed in dollars. Many miners have failed after price falls and halvings.

Summary

Miners add blocks to bitcoin's ledger and are paid a subsidy of new coins plus the fees in the block. The subsidy started at 50 BTC and halves every 210,000 blocks, a schedule that sums to just under 21 million coins, of which 95.8% had been issued by November 2026; the next halving, to 1.5625 BTC, is due around April 2028. A difficulty reset every 2,016 blocks keeps blocks ten minutes apart, so more mining power raises security without raising issuance, and makes mining a contest for a fixed prize. In the three halvings our data covers, bitcoin was higher a year later each time, by less each time, and then fell sharply. That is a record of what happened, with too few cases to say why.

Check your understanding

5 questions on this lesson. Each answer comes with the reason, and nothing is scored.

  1. The total computing power mining bitcoin doubles overnight. What happens to the number of new bitcoin created per day over the following months?

    1. It doubles and stays doubled
    2. It rises briefly, then returns to about 450 a day once difficulty adjusts
    3. It halves
    Show the answer

    B. It rises briefly, then returns to about 450 a day once difficulty adjusts Blocks come faster until the next retarget, when difficulty rises in proportion and blocks return to ten minutes. The subsidy per block does not change.

  2. The last 2,016 blocks took 10 days instead of 14. By how much does difficulty change?

    1. It rises 40%
    2. It rises 28.6%
    3. It falls 28.6%
    Show the answer

    A. It rises 40% New difficulty = old × 14 ÷ 10 = 1.40 times the old level, a 40% rise.

  3. Why does the total supply approach 21 million without the code stating that number?

    1. Miners agreed to stop at 21 million
    2. Each 210,000-block era issues half as many coins as the one before, and that series sums to just under 21 million
    3. The number of blocks is limited to 21 million
    Show the answer

    B. Each 210,000-block era issues half as many coins as the one before, and that series sums to just under 21 million 210,000 × 50 × (1 + ½ + ¼ + …) approaches 21,000,000. Rounding down to whole satoshis makes the exact total 20,999,999.9769.

  4. A miner’s share of network computing power is unchanged, bitcoin’s price is unchanged, and a halving occurs. What happens to its subsidy income in dollars?

    1. It is unchanged, because difficulty adjusts
    2. It halves
    3. It doubles, because coins are scarcer
    Show the answer

    B. It halves The subsidy per block halves, so the same share earns half the coins. Difficulty only helps if other miners switch off.

  5. Bitcoin was higher twelve months after each of the three halvings in the data. What is the soundest conclusion?

    1. Halvings cause price rises
    2. The next halving will be followed by a rise
    3. Three cases, each with other large influences at work, cannot establish cause or predict the next one
    Show the answer

    C. Three cases, each with other large influences at work, cannot establish cause or predict the next one The sample is tiny, the gains shrank each time, each was followed by a fall of a third or more, and the date of every halving was known years ahead.

Sources and further reading

  1. Bitcoin Core source — GetBlockSubsidy in validation.cpp (50 BTC, halved every 210,000 blocks)
  2. Bitcoin Core source — CalculateNextWorkRequired in pow.cpp (the retarget formula and its 4× limits)
  3. Bitcoin Core source — chainparams.cpp (10-minute spacing, two-week retarget window, 210,000-block halving interval)
  4. Bitcoin developer guide — Block chain (difficulty retarget every 2,016 blocks; coinbase outputs spendable after 100 blocks)
  5. Satoshi Nakamoto — Bitcoin: A Peer-to-Peer Electronic Cash System (2008), section 6, Incentive
  6. Blockstream block explorer — timestamps of blocks 210,000, 420,000, 630,000 and 840,000 (the four halvings) and the chain height on November 18, 2026
  7. Finabase chart data — BTC-USD month-end closes, July 2015 to October 2026 (price only), and the bitcoin supply and price shown on finabase.io/crypto on November 18, 2026; schedule and index computed by Finabase