Block Subsidy, Halving & Issuance

Every new bitcoin in existence entered the world the same way: as the block subsidy paid to the miner who found a block. There is no other minting mechanism — no central bank, no pre-mine faucet that keeps dripping. The entire monetary policy of Bitcoin is a few lines of code that halve the subsidy on a fixed schedule, and from that simple rule the famous 21-million cap falls out automatically.

The coinbase: where coins are born

The first transaction in every block is the coinbase transaction (see Coinbase Transactions). It has no real inputs — it conjures value — and it pays the miner two things:

miner reward  =  block subsidy   +   sum of all transaction fees in the block
                 (newly minted)       (recycled from users)

Only the subsidy is new money. The fees are coins that already existed, paid by users for block space (see the fee market). This distinction matters enormously for Bitcoin’s long-run security, which we return to below.

The halving schedule

The subsidy started at 50 BTC per block at genesis and is cut in half every 210,000 blocks — which, at the ~10-minute target block time, is roughly every four years.

Era 1 (blocks 0–209,999)          50    BTC/block
Era 2 (210,000–419,999)           25    BTC/block
Era 3 (420,000–629,999)           12.5  BTC/block
Era 4 (630,000–839,999)            6.25 BTC/block
Era 5 (840,000–1,049,999)          3.125 BTC/block
...                                halving continues until the subsidy rounds to zero

Each of these events is called a halving (or “halvening”). Because the supply increase is known in advance and cannot be altered without breaking consensus, anyone can compute Bitcoin’s exact future issuance for the next century — a stark contrast with discretionary fiat printing.

Why 210,000 blocks, not “every 4 years” directly?

Bitcoin can’t read a calendar — it only counts blocks. The schedule is defined in blocks, so the real-world spacing drifts slightly with actual block times. “Every four years” is an approximation; “every 210,000 blocks” is the law.

Why this sums to 21 million

The total ever issued is just a geometric series. Each era mints 210,000 × subsidy, and the subsidy halves each era:

210,000 × (50 + 25 + 12.5 + 6.25 + ...) BTC
   = 210,000 × 50 × (1 + ½ + ¼ + ⅛ + ...)
   = 210,000 × 50 × 2
   = 21,000,000 BTC   (in the limit)

The infinite sum 1 + ½ + ¼ + … converges to 2, so the whole thing converges to 21 million.

The cap is slightly under 21 million

Because the subsidy is tracked in integer satoshis (1 BTC = 100,000,000 sats) and integer division rounds down, the later, tiny subsidies lose fractions of a satoshi. The true asymptote is a hair below 21,000,000 BTC — and the last new satoshi is expected to be mined around the year 2140, after which the subsidy is zero and miners are paid by fees alone. See The 21 Million Cap & Supply Schedule for the monetary view.

The security-budget question

Here is the rare insight most people miss. Proof of Work’s security comes from the fact that mining is expensive, and miners only spend that money because they’re paid. Today the subsidy dominates that pay. But the subsidy is engineered to vanish:

TODAY:          security paid mostly by  ──►  subsidy (new issuance)
~2140 onward:   security paid entirely by ──► transaction fees

This is an open, genuinely debated question in Bitcoin economics: will fee revenue alone be enough to fund a secure hash rate once the subsidy is gone? The optimistic view is that a thriving fee market (driven by demand for block space) will carry it; the skeptical view worries about a shrinking security budget. Either way, it’s not decided by decree — it’s an emergent property of future demand. We explore both sides in Miner Incentives & the Security Budget.

By the numbers — the four halvings so far

The schedule above isn’t theoretical — four halvings have already fired, each at a height exactly divisible by 210,000:

   Block        Date         Subsidy
   ─────        ────         ───────
   0            Jan 2009     50      BTC   (genesis era begins)
   210,000      Nov 2012     25      BTC
   420,000      Jul 2016     12.5    BTC
   630,000      May 2020     6.25    BTC
   840,000      Apr 2024     3.125   BTC   ← current era
   1,050,000    ~2028        1.5625  BTC   (next, projected)

The dates land roughly four years apart but never on an exact anniversary — because the trigger is a block count, not a calendar (the drift is why “every 210,000 blocks” is the law and “every four years” only the rule of thumb). By the 2024 halving the subsidy had already fallen to one-sixteenth of the original 50 BTC, so new coins now trickle out at a small and ever-shrinking rate — though even that shrunken subsidy still dominates what miners earn, since fees are usually only a few percent of the reward.

The thread

How does this help untrusting strangers agree on one ledger? The issuance schedule is the one piece of “monetary policy” that every participant agrees on without trusting anyone — it’s just arithmetic enforced by every node. No stranger has to believe a central authority about how many coins exist or will exist; they can compute it themselves and reject any block that pays the miner even one satoshi too much. Predictable, verifiable scarcity is itself a form of trustless agreement.

The architect’s lens

Step back from the schedule and answer the five questions that turn an implementer into an architect:

• Why does it exist? To define Bitcoin’s entire monetary policy as a few lines of code anyone can verify — a fixed, geometric issuance (50 BTC halving every 210,000 blocks) that needs no central bank and converges to the 21-million cap.
• What problem does it solve? Trustless, predictable scarcity. No stranger must believe an authority about how many coins exist or will exist; every node computes it and rejects a coinbase that pays even one satoshi too much.
• What are the trade-offs? The disinflation is abrupt (a 50% subsidy cut every ~4 years, not a smooth taper); it’s denominated in blocks, so the calendar drifts; and it deliberately engineers a shrinking security budget — by ~2140 miners are paid by fees alone, a genuinely open question.
• When is this the wrong design? When you need elastic supply — a stablecoin pegged to a fiat unit, or a chain that wants perpetual low inflation (Ethereum’s tail issuance, Monero’s fixed tail emission) to keep funding security indefinitely. Hard scarcity is a feature for sound money, a bug for a unit-of-account that must stay price-stable.
• What breaks if I remove it? Without a capped, halving subsidy there’s no enforceable scarcity — issuance becomes a discretionary knob, the 21M social contract collapses, and the very thing that makes bitcoin behave like “digital gold” disappears.

Check your understanding

1. What are the two components of a miner’s reward, and which one actually creates new bitcoin?
2. State the halving rule precisely — in blocks, not years — and explain why the “four years” figure is only approximate.
3. Show, with the geometric series, why total issuance converges to 21 million.
4. Why is the real supply cap slightly below 21 million?
5. What is the “security budget” problem, and why does the halving schedule make it a long-term question rather than a present-day one?

Show answers

1. The block subsidy (newly minted coins) and the sum of transaction fees (coins recycled from users). Only the subsidy creates new bitcoin; the fees already existed.
2. The subsidy started at 50 BTC and halves every 210,000 blocks. “Four years” is only approximate because Bitcoin counts blocks, not calendar time — at the ~10-minute target 210,000 blocks is roughly four years, but the real-world spacing drifts with actual block times.
3. Each era mints 210,000 × subsidy, halving each era: 210,000 × (50 + 25 + 12.5 + …) = 210,000 × 50 × (1 + ½ + ¼ + …) = 210,000 × 50 × 2 = 21,000,000 BTC, since the infinite sum 1 + ½ + ¼ + … converges to 2.
4. Because the subsidy is tracked in integer satoshis and integer division rounds down, the later tiny subsidies lose fractions of a satoshi — so the true asymptote is a hair below 21,000,000 BTC.
5. Proof-of-Work security comes from mining being expensive, and miners spend that money only because they’re paid — today mostly by the subsidy. But the subsidy is engineered to vanish (~2140), after which security must be funded by fees alone. The halving makes this a long-term open question rather than a present-day one, because today’s subsidy still dominates miner pay.