Bitcoin halving explained: the schedule, the supply maths and what it does not tell you

The bitcoin halving is the point, every 210,000 blocks, at which the number of new bitcoin created in each block is cut in half. It happens roughly every four years and has occurred four times so far, most recently on 20 April 2024, when the reward fell from 6.25 to 3.125 bitcoin per block. It is written into Bitcoin's software, so it is predictable years in advance, and it changes only the flow of new coins, not the coins that already exist.

This guide goes into the mechanics and the arithmetic: how the rule is coded, what each halving did to issuance and supply growth, why the dates drift, what it means for miners, and why the halving is so often misread as a price signal. For the wider picture of how Bitcoin works, read Bitcoin explained first.

Risk first. Bitcoin's price has fallen by more than 70% from previous highs on several occasions. Nothing in this guide is a forecast. If you trade bitcoin through CFDs, leverage magnifies losses, and most retail CFD accounts lose money.

What gets halved, exactly

Every block added to the Bitcoin blockchain contains a special first transaction, called the coinbase transaction, which pays the miner who found the block. That payment has two parts:

• The block subsidy. Brand-new bitcoin created out of nothing. This is the only way new coins enter circulation.
• Transaction fees. The fees attached to every transaction in the block, paid by users. These are existing coins changing hands, not new ones.

The halving applies only to the subsidy. Fees are set by users competing for space in blocks and are untouched by the schedule. When people say "the block reward halved", they mean the subsidy part.

The subsidy started at 50 bitcoin per block when the network launched in January 2009. It has halved four times since, and it will keep halving until it rounds down to nothing.

How the rule is written in the code

The software does not store a calendar of halving dates. It counts blocks. The block height (the number of blocks since the first one) is divided by 210,000, and the result, rounded down, tells the software how many halvings have happened. The starting subsidy, expressed in satoshis, is then halved that many times.

A satoshi is the smallest unit of bitcoin: one hundred-millionth of a coin. The original 50-bitcoin subsidy is 5,000,000,000 satoshis. Because the code works in whole satoshis and simply drops any fraction when it halves, a few things follow:

• Every node on the network calculates the same subsidy for the same block height. There is no vote and no committee deciding the next halving.
• The subsidy can only be a whole number of satoshis. Once halving would produce less than one satoshi, it becomes zero.
• Because fractions are discarded at each step, the total that will ever be issued comes out slightly below 21 million: about 20,999,999.9769 bitcoin.

A miner who tries to pay itself more than the allowed subsidy plus fees produces a block that every other node rejects as invalid. That is what makes the schedule hard to change: it would need the people running nodes, exchanges and businesses to adopt different software, and the existing schedule is one of the properties of Bitcoin its users are least willing to alter.

Every halving so far

Here are the four halvings, with the block height at which each one took effect:

1. 28 November 2012, block 210,000: subsidy cut from 50 to 25 bitcoin.
2. 9 July 2016, block 420,000: 25 to 12.5 bitcoin.
3. 11 May 2020, block 630,000: 12.5 to 6.25 bitcoin.
4. 20 April 2024, block 840,000: 6.25 to 3.125 bitcoin.

The fifth is due at block 1,050,000, when the subsidy would fall to 1.5625 bitcoin. It is widely expected in 2028, but the exact day depends on how quickly blocks are found between now and then. Some sources list the first halving as 27 November 2012; the difference is a time zone, since the block arrived late on 28 November UTC.

Why the dates drift

The four-year rule of thumb comes from the target of one block every ten minutes: 210,000 blocks multiplied by ten minutes is about 1,458 days, or just under four years. In practice the gaps between halvings have not been exactly four years, for two reasons.

Difficulty only adjusts every 2,016 blocks. The network sets the difficulty of the mining puzzle so that, on average, blocks arrive every ten minutes. It recalculates the difficulty every 2,016 blocks, roughly every two weeks, based on how fast the previous 2,016 blocks were found. Between adjustments, if more computing power joins, blocks come faster than ten minutes; if miners switch off, slower.

Computing power has mostly grown. Over Bitcoin's history the total computing power on the network, called the hash rate, has risen enormously, with occasional sharp drops. Because difficulty always lags, a rising hash rate means blocks arrive a little faster than target on average, which pulls halvings slightly earlier than a strict four-year count would suggest. A sudden drop in hash rate, such as when large numbers of miners have had to relocate, does the opposite until the next adjustment catches up.

So if you see a countdown to the next halving, treat the date as an estimate based on recent block times. The block height is the fixed part.

The supply maths

Each era between halvings lasts 210,000 blocks, so the number of coins issued in an era is simply 210,000 multiplied by that era's subsidy. Adding them up shows how front-loaded Bitcoin's issuance is.

Coins issued in each era

• Era 1 (2009 to 2012), 50 per block: 10,500,000 bitcoin. Cumulative: 10,500,000, or half of everything that will ever exist.
• Era 2 (2012 to 2016), 25 per block: 5,250,000. Cumulative: 15,750,000.
• Era 3 (2016 to 2020), 12.5 per block: 2,625,000. Cumulative: 18,375,000.
• Era 4 (2020 to 2024), 6.25 per block: 1,312,500. Cumulative: 19,687,500.
• Era 5 (2024 to about 2028), 3.125 per block: 656,250. Cumulative by block 1,050,000: 20,343,750.

These are the scheduled figures. The real number of coins in circulation is very slightly lower, for technical reasons such as the subsidy of the very first block, which cannot be spent, and a small number of blocks in which miners claimed less than they were allowed.

Each era issues exactly half as many coins as the one before. That is a geometric series: the remaining eras together can never add more than the current era has just issued. After block 840,000, all future issuance combined is about 1.31 million bitcoin, the same as era 4 alone.

Worked example: the 20 millionth bitcoin

Using the scheduled figures, here is a quick illustration of how to work out when a supply milestone falls.

1. Supply at block 840,000 (the April 2024 halving): 19,687,500.
2. Coins still needed to reach 20,000,000: 312,500.
3. At 3.125 bitcoin per block: 312,500 ÷ 3.125 = 100,000 blocks.
4. At about 144 blocks a day: 100,000 ÷ 144 ≈ 694 days, or a little under two years.

That puts the milestone around block 940,000, which places it in the early months of 2026. Running numbers like these yourself is a good habit: claims about Bitcoin's supply are easy to check, because the schedule is public and mechanical.

Daily and annual issuance

Assuming exactly 144 blocks a day (six an hour), daily issuance in each era was:

• 50 per block: 7,200 bitcoin a day.
• 25 per block: 3,600 a day.
• 12.5 per block: 1,800 a day.
• 6.25 per block: 900 a day.
• 3.125 per block: 450 a day, or roughly 164,000 a year.

Supply growth, as a percentage

What usually matters more than the raw number of coins is how fast supply is growing relative to what already exists. An illustrative calculation around the 2024 halving:

• Just before it: about 328,000 new coins a year on a supply of about 19.69 million, which is supply growth of roughly 1.7% a year.
• Just after it: about 164,000 a year on the same supply, roughly 0.8% a year.

Each halving roughly halves the growth rate, and because the existing supply keeps rising, the percentage falls a little further during each era. Bear in mind this is a mechanical number, not an inflation rate in the consumer-price sense. It tells you about new supply, not about the purchasing power of a coin.

When does it end?

The subsidy keeps halving until it rounds down to zero. Starting from 5,000,000,000 satoshis, the 32nd halving leaves a subsidy of exactly one satoshi per block, and the 33rd takes it to zero. The 33rd halving falls at block 6,930,000, which at ten minutes per block is about 132 years after the network started in 2009, landing around the year 2140. That is where the commonly quoted "last bitcoin in 2140" comes from.

Two points are worth keeping in perspective:

• By the 2030s the subsidy will already be small. After the seventh halving, around 2036, it would be under 0.4 bitcoin per block. Most of the long tail is tiny amounts.
• The 21 million figure is a ceiling on issuance, not on usable coins. Some bitcoin has been permanently lost through forgotten keys and discarded devices. Estimates of how much vary a great deal, so be sceptical of any precise number.

What the halving means for miners

Miners are businesses. Their revenue is the subsidy plus fees, paid in bitcoin, and their costs are mostly electricity, hardware and premises, paid in ordinary currency. A halving cuts the subsidy part of revenue in half overnight, while costs stay the same.

Worked example: a miner's revenue per block

An illustrative calculation with round numbers, not real market data:

• Suppose fees average 0.25 bitcoin per block.
• Before the 2024 halving: 6.25 + 0.25 = 6.5 bitcoin per block.
• After it: 3.125 + 0.25 = 3.375 bitcoin per block.
• Revenue per block falls by about 48%, slightly less than half, because fees were unaffected.

If the miner's share of the network and its costs stay the same, and the bitcoin price does not change, its revenue in currency roughly halves. Whether it stays profitable then depends on its electricity price and how efficient its machines are.

How the network absorbs it

After a halving, the least efficient miners may switch off their machines. If enough do, blocks slow down until the next difficulty adjustment lowers the difficulty, which makes mining a little cheaper for those who remain. The system is designed to rebalance itself this way, and so far it has: the network has kept producing blocks through every halving.

Some observers watch miner behaviour around halvings, for example whether miners sell more of their coins to cover costs. These flows are hard to measure precisely and their effect on price is debated. Our guide to on-chain metrics explains what can and cannot be read from blockchain data.

The long-term question: fees

As the subsidy shrinks, transaction fees have to make up a growing share of what miners earn if the network's security spending is to hold up. Whether fees will be enough in a few decades is one of the genuinely open questions about Bitcoin, discussed seriously by researchers and developers. There is no settled answer, and you should be wary of anyone who presents one with certainty. The comparison of proof of work and proof of stake covers why security spending matters at all.

The halving and the price

This is where most of the confusion lies, so it is worth being precise about what can and cannot be said.

The supply argument

The argument for the halving mattering runs like this: if demand stays the same and the number of new coins that miners need to sell is cut in half, there is less selling pressure. Over time, a lower flow of new supply could matter to price.

The counter-arguments

• It is known in advance. The halving schedule has been public since 2009. Markets generally try to price in information that everyone already has. A predictable event is not a surprise.
• New supply is a small share of trading. Daily issuance after 2024 is about 450 coins. The amount of bitcoin that changes hands on exchanges and in derivatives markets each day is many times larger. Changes in demand, leverage and sentiment can easily swamp a change in miner supply.
• Four data points. There have been four halvings. Each happened in very different conditions: different interest rates, different regulation, different market access (for example, spot bitcoin exchange-traded funds began trading in the US in January 2024, a few months before the fourth halving). Four observations cannot establish a reliable pattern, and attributing what followed each halving to the halving itself is guesswork.
• Price has moved both ways. History shows sharp rises and deep falls in the months and years around halvings. Our guide to crypto market cycles explains why past cycles prove nothing about the next one.

Stock-to-flow and similar models

You will come across models that predict price from scarcity, the best known being stock-to-flow, which compares existing supply (the stock) with annual new supply (the flow). Because halvings change the flow on a fixed schedule, these models produce precise-looking price paths. Treat them with caution. They are fitted to a short history, they ignore demand entirely, and predictions based on them have missed badly. A model that only looks at supply cannot tell you what people will pay.

The honest summary is that the halving is a certain change in supply growth with an uncertain effect on price.

Trading around a halving

The halving itself does not change how a bitcoin CFD works, but the period around one tends to attract attention, commentary and, at times, heavy trading. If you trade bitcoin through CFDs, the usual rules apply with extra force.

• Volatility can rise. Large moves in either direction are possible when many traders are focused on the same date. Read crypto volatility and risk for how these moves reach an account.
• Leverage is capped for a reason. Under ESMA rules, retail clients in the EU can use at most 2:1 leverage on crypto CFDs, so you must put up at least 50% of the position's value as margin.
• Size from your stop, not from the story. Decide where you would be wrong, then size so that a loss at that point is a small part of your account.

Worked example: sizing a bitcoin CFD

An illustrative example; the price is a round number, not a quote.

• Account: €10,000. Maximum risk on the trade: 1%, or €100.
• Bitcoin CFD price: €50,000. You place a stop 8% away, at €46,000, so the stop distance is €4,000 per bitcoin.
• Position size: €100 ÷ €4,000 = 0.025 bitcoin.
• Exposure (notional value): 0.025 × €50,000 = €1,250.
• Margin at 2:1: €1,250 × 50% = €625.

If the stop is hit you lose about €100 plus costs, and possibly more if the price gaps through the stop, which happens in crypto, including on weekends when CFD pricing may differ from exchange trading hours. The crypto position size calculator does this arithmetic for you, and the crypto margin calculator shows the margin required. Remember that holding a CFD position overnight carries financing charges, explained in crypto CFD overnight fees.

Common mistakes

• Treating the halving as a buy signal. It is a known supply change. Its effect on price, if any, cannot be isolated from everything else happening at the time.
• Confusing the subsidy with the total reward. Fees are not halved. In periods of heavy network use, fees can be a meaningful share of what miners earn.
• Relying on a countdown date. The halving happens at a block height. The calendar date is an estimate that can move by days or weeks.
• Thinking existing coins are affected. Nothing happens to coins already issued. Your bitcoin does not change on halving day.
• Trusting precise-looking models. Supply-only models ignore demand and have a poor record. Precision is not accuracy.
• Over-leveraging into a well-publicised event. The more people watch a date, the more crowded positioning can become, and crowded positions can unwind violently. See why most retail CFD traders lose.

Frequently asked questions

What is the bitcoin halving in simple terms?

It is the scheduled cut, every 210,000 blocks, in the number of new bitcoin paid to miners for each block. The cut is 50% each time. It slows the creation of new coins and has happened four times, in 2012, 2016, 2020 and 2024.

When is the next bitcoin halving?

At block 1,050,000, when the subsidy will fall from 3.125 to 1.5625 bitcoin. That is expected in 2028, but the exact date depends on how quickly blocks are found, so countdown dates are estimates.

How many bitcoin halvings will there be?

The subsidy reaches zero after the 33rd halving, around the year 2140. The last non-zero subsidy is a single satoshi per block. After that, miners would be paid by transaction fees alone.

Why is it 210,000 blocks?

It is the parameter chosen in Bitcoin's original software. At the target of one block every ten minutes, 210,000 blocks is just under four years. Combined with a 50-bitcoin starting subsidy, it produces a total issuance of just under 21 million coins.

Can the halving schedule be changed?

Technically, only if the people running Bitcoin's software adopted a new version with different rules, and blocks created under the old and new rules would no longer be compatible. The fixed supply schedule is one of the features Bitcoin's users have shown the least appetite to change, so most participants treat it as permanent.

Does the halving affect bitcoin CFDs?

Not mechanically: a CFD tracks the price, and the contract terms do not change at a halving. Any effect comes through price movement and volatility around the event. Leverage on crypto CFDs is limited to 2:1 for EU retail clients, and losses can exceed what you expected if the price gaps through a stop. Our comparison of crypto CFDs and buying crypto explains the differences.