Fixed supply vs flexible monetary policy: why Bitcoin's cap holds

Most people know Bitcoin has a 21 million coin limit. Fewer appreciate the part that actually matters: that limit has never been changed, and the reason it won’t be has nothing to do with the code being unchangeable. Anyone can edit Bitcoin’s software. What can’t be edited is the incentive structure — every holder loses if the supply inflates, so any attempt mobilises the entire network against it. Most other crypto assets work differently: their issuance can be, and has been, adjusted through governance. For anyone storing value or building financial products on top, the question isn’t “what’s the supply?” — it’s “can this be diluted by whoever holds governance power?”

Does Bitcoin really have a fixed supply?

Yes, and it’s enforced from the bottom up. Every full node — a computer storing the whole blockchain and checking every transaction — independently verifies that each new block creates no more Bitcoin than the schedule allows. Any block that over-issues is simply rejected.¹ The schedule itself is mechanical: the block reward (what proof-of-work mining pays out) halves roughly every four years, from 50 BTC in 2009 to 3.125 BTC since April 2024, trailing to zero around 2140. The headline number is 21 million, though the effective total is closer to 20,999,950 — a quirk in the genesis block left its original 50 BTC permanently unspendable.² No committee sets this. No vote adjusts it. It’s arithmetic that thousands of independent machines enforce in parallel.

Why “never been changed” matters more than “fixed”

“Fixed” makes it sound like the cap is physically locked. It isn’t — Bitcoin is open-source software, and anyone can write a version that mints more. What makes the cap credible is that no such version would ever be adopted. Changing the monetary rule requires a hard fork³ accepted by the overwhelming majority of Bitcoin’s roughly 18,500–24,800 reachable nodes,⁴ and every one of those operators — like every holder — would be voting to devalue their own coins. The incentives run powerfully in one direction. (This is how Bitcoin handles every rule change: node consensus, not committee.)

We have a real-world stress test for it. During the blocksize wars of 2015–2017, a coalition of major exchanges, miners, and companies representing more than half of network activity tried to push through a comparatively minor change — a block-size increase — and failed.⁵ If that coalition couldn’t move a parameter about transaction throughput, the odds of anyone changing the core supply rule are effectively nil. This is why Bitcoin’s monetary policy is described as credibly neutral: not because change is technically impossible, but because no party can benefit from it without provoking a larger, more motivated coalition against them.

What is monetary policy in crypto, and can it change?

For most other networks, yes — routinely. Ethereum has revised its monetary policy several times through governance: EIP-1559 in 2021 began burning ETH on every transaction, and the 2022 Merge ended mining rewards and cut daily issuance by around 88%, leaving supply oscillating between mildly inflationary and deflationary depending on usage.⁶ Solana and Cardano likewise tune issuance through governance. None of this is necessarily mismanagement — but it requires only a relatively small set of decision-makers: core developers, a foundation, large staking providers, validators. That’s a very different thing from needing a globally distributed base of independent nodes to agree.

The distinction is about who holds the lever, not whether the policy can change at all. On Ethereum, monetary policy is a governance outcome. On Bitcoin, it’s a near-immovable constraint. Which you prefer depends entirely on how much you want to trust the people holding the lever.

Why this matters for savers and builders

For a saver, this is the whole question. If you’re holding an asset for a decade, the thing that should keep you up at night isn’t volatility — it’s dilution. Can parties with governance power expand the supply and quietly transfer value away from you? On Bitcoin, the answer is no, in a way backed by the strongest incentive alignment in the system. On most other networks, the answer is “it depends on the governance process.”

For a builder, it’s the same question wearing a different hat. The case for Bitcoin as a long-term store of value rests on scarcity you can actually rely on, and financial products — lending, savings, collateralised structures — need that reliability as a foundation. A scarcity guarantee a committee can revise is not much of a guarantee. One that requires the entire network to act against its own interest is something you can build on.

Common questions

Can Bitcoin’s 21 million limit ever be changed? In principle anyone can write software that mints more, but it would only take effect if the vast majority of nodes adopted it — and since that would devalue everyone’s holdings, the incentives make it effectively impossible. The limit has never been changed.

Why is the cap “21 million” when the real number is lower? A quirk in the original code left the very first block’s 50 BTC permanently unspendable, so the effective maximum is about 20,999,950 BTC. The difference is tiny and doesn’t affect the scarcity argument.

Has any other cryptocurrency changed its monetary policy? Yes — Ethereum has changed its issuance multiple times (a fee burn in 2021, the end of mining rewards in 2022), and networks like Solana and Cardano adjust issuance through governance. This is normal for governed chains; it’s the thing Bitcoin specifically avoids.

What happens when all 21 million are mined? New issuance trails to zero around 2140, after which miners are paid through transaction fees rather than newly created coins. The supply stops growing; the network keeps running.

Footnotes

1. Bitcoin Wiki, “Controlled supply.” https://en.bitcoin.it/wiki/Controlled_supply ↩

2. A quirk in Bitcoin’s original code left the genesis block’s 50 BTC permanently unspendable, making the effective maximum supply 20,999,950 BTC rather than 21,000,000. Bitcoin Wiki, “Genesis block.” ↩

3. Hard fork — a protocol change that makes old software incompatible with new, requiring near-universal adoption to avoid splitting the chain. Investopedia, “Hard Fork.” https://www.investopedia.com/terms/h/hard-fork.asp ↩

4. Bitnodes — approximately 18,500–24,800 reachable full nodes across 181 countries, 64.58% via Tor. https://bitnodes.io/ ↩

5. Jonathan Bier, The Blocksize War (2021) — the 2015–2017 dispute in which a coalition of major exchanges, miners, and companies failed to force a block-size increase. ↩

6. Ethereum’s EIP-1559 (August 2021) introduced a fee burn; the September 2022 Merge ended proof-of-work issuance, cutting daily issuance by ~88%. Ethereum.org, “The Merge.” https://ethereum.org/en/roadmap/merge/ ↩