A hyperstructure is, in Jacob Horne's 2022 definition, a crypto protocol that "can run for free and forever, without maintenance, interruption or intermediaries". The term names a class of on-chain infrastructure rather than a specific technology: any protocol meeting the criteria below qualifies, and — as the essay is careful to stress — being built on a blockchain is not by itself sufficient.

The essay has become one of the standard references for the argument that a protocol need not extract fees from its users in order to be worth owning, and it supplies much of the vocabulary now used when a DAO debates whether to switch on a protocol fee.

The seven properties

Horne lists seven characteristics, all of which a protocol must have to count as a hyperstructure:

• Unstoppable — "the protocol cannot be stopped by anyone. It runs for as long as the underlying blockchain exists."
• Free — "there is a 0% protocol wide fee and runs exactly at gas cost."
• Valuable — "accrues value which is accessible and exitable by the owners."
• Expansive — "there are built-in incentives for participants in the protocol."
• Permissionless — "universally accessible and censorship resistant. Builders and users cannot be deplatformed."
• Positive sum — "it creates a win-win environment for participants to utilize the same infrastructure."
• Credibly neutral — "the protocol is user-agnostic."

The properties are presented as consequences of one another rather than as an arbitrary list. Unstoppability is the root: because a deployed contract needs no operator, labour, or subsidy to keep running, there is no ongoing cost that a protocol fee has to cover, which is what makes "free" affordable in the first place. Permissionlessness and credible neutrality then remove the reasons anyone would rebuild the same thing elsewhere, which is what makes the environment positive-sum. On the last point Horne explicitly follows Vitalik Buterin's 2020 argument for credible neutrality, also discussed under governance minimization.

The threat of the fee

The essay's most contested claim is that a protocol can be "simultaneously free forever to utilize and extremely valuable to own." Ownership, in this account, is not a claim on revenue but the control of a fee switch that has not been flipped — what Horne calls "the threat of the fee".

The argument is that actually turning the fee on is value-destructive, because it "would immediately lead to an incentivized fork, since there's now a clear reason for new entrants to do it themselves." The right to switch it on is therefore analogous to an NFT owner's right to burn: a rational owner will not exercise it, but the fact that they could is what makes the position worth something. Horne treats the right to sell or transfer the switch as the second ownership right, since that is what lets a market put a price on it. The wider claim drawn from this is that "for-profit" is a "skeuomorphic mode of operation" — that value creation and profit extraction have come apart, and a protocol can be built "for-public" instead.

The canonical worked example was the Uniswap fee switch, and it is no longer a threat: Uniswap governance flipped it. The UNIfication proposal — posted on-chain on 18 December 2025 by Uniswap Labs and the Uniswap Foundation, and explicit that "the Uniswap protocol includes a fee switch that can only be turned on by a UNI governance vote" — asked holders to turn protocol fees on and route them into a UNI burn. It passed and executed on 28 December 2025; the tally and the burn mechanics are set out on fee switches and value accrual.

That converts the essay's most contested claim from an argument into a testable one, because Horne's case for never exercising the switch was a prediction about what happens if you do: turning the fee on "would immediately lead to an incentivized fork." Uniswap's own February 2026 report on the rollout says the opposite happened on the measure it chose — "market-adjusted TVL up on Ethereum mainnet since December" — and it is the interested party's own reading, so weigh it as such. The unarguable part is the revealed preference of the electorate. Two months after activation the same holders were asked to extend the fee to eight further chains and to every remaining mainnet v3 pool, and the Snapshot record closes at 35,153,339 UNI for, 0 against, 0 abstaining across 150 votes (re-queried 3 September 2026). A DAO that had been forked for switching the fee on does not unanimously vote to switch it on in more places.

None of this refutes the hyperstructure criteria, and it is worth being precise about what it does. "Free" is a definitional test, not a prophecy: a protocol that turns its fee on simply stops being a hyperstructure from that day, which is exactly what the essay says should happen. What the episode falsifies is the weaker empirical claim carried alongside it — that the threat is more valuable unexercised than the revenue is exercised, because exercising it triggers the fork. On the largest case anyone had, the holders disagreed and the fork did not arrive on the schedule the argument predicted.

Expansive fees

"Free" applies to the protocol-wide fee, not to every payment inside the system. Hyperstructures are expected to carry expansive fees: incentives "utilized by anyone adding value in a codified manner on top of the protocol," open to the whole ecosystem rather than accruing to the protocol's operator.

Horne's worked example is the Uniswap liquidity-provider fee, which pays anyone supplying the resource the protocol needs — liquidity — and "is not paid to Uniswap"; the second is Zora's finder's fee, which pays whoever brings the eventual buyer. The general rule the essay draws is that in a hyperstructure "any participant's value captured is synonymous with their value generated," and that a system permitting any party to extract value without creating it is not a hyperstructure.

What does not qualify

The essay spends as much effort on exclusions as on the definition, since most of what is colloquially called a protocol fails at least one criterion. Two of its own counterexamples:

• Dependence on off-chain infrastructure. The Wyvern protocol used by OpenSea "cannot operate without offchain orders being maintained in a privately operated database" — so if the platform goes down, the market goes with it. Being protocol-based is not enough; a hyperstructure is "entirely onchain."
• Upgradeable admin control. A fully on-chain protocol whose administrators can "completely upgrade and change the rules" is, in the essay's terms, still a platform rather than a hyperstructure — the same reasoning that makes immutability central to governance minimization.

Horne is explicit that these are not criticisms: such systems "are just not Hyperstructures." The distinction matters mainly because the label carries a promise to builders — that anything built on top cannot later be deplatformed or repriced.

How Caper approaches this

The hyperstructure criteria are a useful yardstick precisely because most protocols, capers included, do not clear all seven. Caper charges protocol fees rather than running at gas cost, so it is not "free" in Horne's strict sense; the fee levers and where the proceeds go are set out on bonding curve and raising funds.

Where the frames do overlap is on the properties Horne roots the rest in. Caper's contracts are deployed on Radix and its market is permissionless and always-on, so a caper's holders are not relying on an operator to stay in business in order to trade or govern; and the essay's "valuable" criterion — value that is "accessible and exitable by the owners" — is close to the reasoning behind a caper's standing exit right. The honest reading is that Caper adopts the exit and permissionlessness arguments while rejecting the zero-fee one.

References

• Jacob Horne, Hyperstructures (jacob.energy, 16 January 2022) — the source of the definition and all quotations above.
• Related reading on this wiki: governance minimization, fee switches and value accrual, progressive decentralization.