Seigniorage

The monetary concept, before anyone tokenised it

Central banks define this plainly. The European Central Bank describes seigniorage as the profit made from issuing currency, arising because the face value of money exceeds what it costs to produce.¹ A high-denomination note costs pennies to print and enters circulation at face value.

The formal treatment matters more than the anecdote, because it puts seigniorage inside a budget constraint. Obstfeld derives the revenue-maximising rate of monetary growth: past a certain rate of issuance, faster issuance yields the issuer less real revenue, because holders adjust.² Anyone who has watched a project answer a falling token price by raising emissions has watched that result being rediscovered without the maths.

What it becomes when the issuer is a protocol

Mint a token and the marginal production cost is a transaction fee, while the unit enters circulation at market value. The difference is seigniorage, and unlike a central bank the protocol usually hands it straight to somebody else: a validator, a liquidity provider, an airdrop recipient. The issuer captures nothing. The recipient captures it, and existing holders fund it.

That reframes an argument which usually gets stuck. Rather than debating whether emissions are inflationary, ask who receives the seigniorage and what the protocol bought with it. A validator reward buys security. A liquidity incentive buys depth for as long as you keep paying. An unlock buys nothing, because those tokens were already promised. If what the protocol receives is worth less than the value transferred out of existing holders, the emission is a transfer with a story attached.

The seigniorage-share stablecoin design

The design class carrying the name works like this. One token is meant to hold a stable value. A second absorbs the volatility, and the peg is defended by a two-way exchange between them: below target, holders are given an incentive to destroy the stable unit in return for newly minted units of the volatile one; above target, the reverse. The volatile token is the residual claim, and it is the seigniorage instrument.

The arbitrage described here is reconstructed from independent academic postmortems rather than from the issuer's own materials, which we could not verify at a citable URL. Worth saying out loud, because the marketed version and the analysed version of a failed mechanism are not always the same document.

The documented failure, and what it demonstrated

Terra's UST and LUNA is the case with real academic coverage. Harald Uhlig's NBER working paper analyses the collapse formally, as a run rather than an isolated exploit.³

The mechanism that failed is worth stating precisely. The absorbing token's capacity to defend the peg is bounded by its own market value, and that value is a claim on the same system. When the peg comes under real pressure, the collateral defending it reprices downward at the same time as the mechanism issues more of it. The design assumes the absorber has capacity in the exact scenario where it does not.

Our reading, offered as interpretation rather than as a finding in that paper, is that this is not a parameter failure. No mint rate, cap or cooldown makes an absorbing asset uncorrelated with the thing it absorbs. That is a structural property of using your own equity as reserves, and it is why the two-token separation rule is a hard constraint rather than a preference.

What a founder should take from it

Two things, and neither requires believing that every algorithmic design is doomed. Name your seigniorage recipients and price what the protocol receives from each, because that is how you tell a purchase from a transfer. And where any mechanism depends on minting an asset under stress, model the correlation between that asset's value and the stress itself rather than assuming a static reserve. External, uncorrelated backing is the expensive option, and the cheap alternative is cheap because it is a claim on yourself. Nothing here is a recommendation about any asset.