Tokenomics: supply, allocation, emission, and sinks.

The four questions

Tokenomics has become a word that means almost anything, which makes it hard to evaluate. It is more tractable if you reduce it to the four decisions that every design has to make, in this order, because each one constrains the next.

Supply design

Three broad patterns exist, and most real designs are a hybrid. A fixed supply mints everything at genesis and distributes it over time from escrow, so total supply never changes and only circulating supply moves. A schedule-based emission mints new tokens continuously, often to pay for security or to incentivise participation. An open-ended design has no maximum at all, which means fully diluted valuation is undefined and comparisons using it are meaningless.

A published hybrid makes the pattern concrete. Uniswap minted one billion UNI at genesis and stated a perpetual inflation rate of 2% per year beginning after four years. For the first four years the design behaves like a fixed supply with unlocks; after that it behaves like an inflating one. Anyone modelling long-run supply has to handle both regimes, and anyone quoting a fully diluted figure has to say which regime they assumed.

Burns run the other way. A protocol that destroys tokens — using fee revenue, or as part of a transaction mechanic — removes them from total supply permanently. Data providers subtract verifiably burned tokens from total supply, so a burn changes the published denominator rather than just moving tokens to an inert address. Whether the burn is meaningful depends entirely on its size relative to emissions, which is an arithmetic question rather than a narrative one.

Allocation: what three real launches actually did

This is where invented conventions do the most damage. It is common to encounter a confident claim that the standard split is some particular set of percentages. There is no such standard, and the fastest way to see it is to lay three well-documented launches side by side using only their own published figures.

The investor line clusters — roughly 17% to 18% in all three. Almost nothing else does. Contributor allocations differ by nearly eight percentage points. Optimism carved out a fifth of supply specifically for retroactive public goods funding, a category the other two do not have. Arbitrum put over a third into a DAO treasury. Uniswap made a claimable distribution to prior users its single largest immediate move and then handed 43% of supply to a governance treasury.

Those are not three implementations of one convention. They are three different theories about who a network belongs to, expressed in numbers. Copying percentages from one of them without adopting its governance model gives you the arithmetic without the reasoning.

Emission, inflation, and who pays for it

Where a design emits new supply, someone is paying for something. Proof-of-stake networks emit to compensate validators for securing the chain. Applications emit to bootstrap liquidity or usage. In both cases the emission is a real cost borne by existing holders through dilution, and the honest way to evaluate it is to ask what is being bought and whether it persists once the emission stops.

• Emission rate. Expressed as a percentage of supply per year, it is directly comparable across designs; expressed as tokens per block, it is not.
• Recipients. Validators, liquidity providers, and users are different constituencies with different behaviour when incentives end.
• Duration. A fixed-term emission programme and a perpetual one are different commitments, and the difference should be legible in the documentation.
• Governance control. Whether the rate can be changed by a vote, and by whom, is often more consequential than the initial rate.
• Offsetting sinks. Emission net of burns is the number that matters for float; emission alone is half a sentence.

Sinks and utility

A sink is any mechanism that removes tokens from circulation or immobilises them. Burns destroy supply outright. Staking locks it for a period, often with a withdrawal delay. Collateral requirements tie it up while a position is open. Fee payment routes it somewhere it does not immediately return from. Governance participation may require locking for a term.

The reason sinks belong in a discussion of cap tables rather than a discussion of markets is that they change what circulating supply means. Tokens locked in a staking contract are in existence and are owned by someone, but they are not available to trade. Whether a supply definition counts them is a judgement call, providers make it differently, and a project that does not publish its own definition is letting other people answer the question.

It is worth being careful about what utility claims can support. A sink is a supply mechanic. It is not a statement about value, and nothing on this page should be read as suggesting any design produces a particular outcome for a token’s price. Supply design determines float; float is one input among many into a market you cannot model from a documentation page.

From allocation to the register

Every tokenomics decision lands as a structure on the token cap table, and designs that look elegant in a diagram frequently produce records that are painful to maintain.

The last thing worth saying is that tokenomics is usually presented as a launch decision and is really a governance decision with a long tail. The genesis allocation is fixed the moment the token generation event happens. Everything after that — emission changes, treasury spending, new incentive programmes — runs through whatever governance the allocation itself created. That is why the allocation table and the governance design have to be read together, and why /equity/token-cap-table treats the register as a living document rather than a launch artefact.