Ethereum does not have a hard cap on supply, but it *does* have a hard cap on its inflation rate

Supply debates often get stuck in a false binary: either an asset has a hard cap like Bitcoin's 21M, or it is "uncapped" in a way that implies runaway dilution. Ethereum does not fit neatly into either bucket.

A more accurate post-Merge description is: no fixed terminal cap, but an issuance model that allows an upper bound on supply growth over any chosen time window, under stated assumptions.

The goal here is not to relabel ETH as "capped". It is to tighten the language so people can argue about the same object.

Definitions

A lot of confusion comes from using "capped" to mean different things: total supply at time T, the annualized issuance rate, or net issuance after burn.

What changed after the Merge (and why it matters)

Since the Merge, Ethereum's block production and consensus rewards are governed by proof-of-stake (PoS) Beacon Chain specs, not proof-of-work mining economics. In PoS, issuance follows protocol reward formulas and is tied to staking participation rather than hashpower competition.

That makes Ethereum's baseline issuance path rule-based rather than what you see with fiat or centralized currencies, where it is discretionary on a per-period basis.

"Issuance cannot go to literal zero" — the security-budget tradeoff

A common intuition is that if a network wants non-zero security, it needs an ongoing payment stream to validators, which may imply ongoing issuance if fees alone are not expected to cover it. In practice, Ethereum validator revenue can include issuance, transaction tips/priority fees, and MEV (maximal extractable value). The base fee is burned rather than paid to validators.

Ethereum trades the simplicity of a supply hard-cap for the security of an incentive design that can keep validator participation viable across different fee/usage patterns.

A "finite supply on any finite timeline"

Even without a hard cap, it is possible to **bound the maximum supply increase over a specified interval** by choosing conservative (worst-case) assumptions, such as:

• assume minimal burn (including the conservative bound burn = 0), and

• assume maximal issuance consistent with the staking-dependent issuance formula.

Under these assumptions, an upper bound for supply at time T can be computed for Ethereum: not a guaranteed endpoint, but a ceiling.

This ceiling is often calculated at approximately 1.51% annualized gross issuance¹ . This figure represents a theoretical maximum inflation rate that assumes 100% of all ETH is staked and zero fees are burned — a scenario that functions as a "worst-case" bound for dilution.

This can be characterized as a supply cap that adjusts upward at a slow and predetermined rate. The "cap" is not a fixed terminal number, but a deterministic upper-bound curve that rises over time under the protocol's rules.

Sources

¹ Edgington, Ben. "Issuance." Upgrading Ethereum. Retrieved from https://eth2book.info/latest/part2/incentives/issuance/. (Provides the technical derivation for validator rewards, demonstrating that issuance scales sub-linearly with participation, capping out near ~1.5–1.7% even at theoretical maximum staking levels). ² "ETH: Total Supply Through the History of Ethereum Updates." Glassnode. Retrieved from https://studio.glassnode.com/metrics?a=ETH&m=supply.Current.