Velocity becomes a price problem when spendable float outruns demand
Token price instability is often a velocity problem disguised as a demand problem. Velocity is the frequency with which one unit of money is used in transactions, but in crypto the relevant variable is not the turnover of total supply. It is the turnover of the supply that is actually available to clear payments, arbitrage, and exits.
The useful starting identity is the crypto exchange-rate equation. BIS shows fiat price per coin as transaction demand in dollars divided by the product of payment-available coins and their payment velocity, which means price falls when either economically spendable supply rises or each coin turns over faster.
That framing explains why thin transactional float can produce violent price moves. BIS argues that the “crypto multiplier” rises when a larger share of coins is held as a store of value rather than used for payments, and that high volatility is likely to persist unless use shifts toward actual payment activity.
Instability also feeds back into utility. A 2024 empirical study of 58 tokens found that tokens used as a medium of exchange were negatively correlated with price volatility, while a 10% increase in volatility was associated with a 3.96% to 5.88% decline in active addresses. Price instability does not only hurt holders. It can reduce network usage itself.
Model effective velocity, not headline turnover
Headline turnover is a bad proxy for economic velocity. Coin Metrics defines adjusted transfer value as transferred value after removing noise and artifacts, including early spends, self-churn, cold-wallet shuffles, and other non-economic activity. Glassnode defines velocity as on-chain transaction volume divided by market cap, and Coin Metrics explicitly describes NVT as the opposite of velocity. Those definitions are only useful if the numerator is cleaned and the denominator reflects actually tradable supply.
Free float matters more than naive circulating supply. Adjusted free-float calculations exclude foundation holdings, team holdings, vesting balances, some governance-staked tokens, burned or provably lost tokens, and very old inactive coins because those balances are not realistically available to the market.
Aggregate velocity is also too coarse for design decisions. The MicroVelocity paper finds that velocity is highly heterogeneous across agents and that high-velocity intermediaries can dominate observed aggregate velocity. If one exchange, bridge, or rewards recycler is turning the token constantly, the headline ratio can look active while user-level holding behavior barely changes.
| Model component | Why it matters | Minimum treatment |
|---|---|---|
| Adjusted transfer value | Raw on-chain volume overstates real economic activity. | Use adjusted transfer value instead of gross transfer volume. |
| Spendable float | Team, vesting, burned, lost, and some escrowed balances are not equally tradable. | Build a free-float series and refresh it around unlock events. |
| Cohort behavior | Intermediaries can dominate aggregate velocity. | Segment turnover by cohort, venue, or wallet class where possible. |
| Time dimension | Old dormant supply re-entering circulation is different from routine churn. | Track coin days destroyed or an equivalent age-spend metric. |
| Wrapper leakage | Locked base assets can regain liquidity through wrappers. | Net hard locks against liquid claims that re-enter circulation. |
Build a state-based velocity model
A practical token velocity model should be state-based, not scalar. Start by separating token stock into five buckets that map to core token economy design components: permanently removed supply, hard-locked strategic supply, conditionally locked supply, synthetic liquid claims, and freely spendable float. The model should calculate economic velocity only on the last bucket, while scenario tests describe how tokens migrate between buckets over time.
Effective Velocity_t = Adjusted Economic Transfer Volume_t / Spendable Float_t. Spendable Float_t should usually be modeled as on-chain circulating supply minus burned, lost, team, foundation, vesting, and genuine governance escrows, plus any liquid wrappers or borrowable claims that restore transferability in practice. That formula is an analytical inference, but it follows directly from the BIS emphasis on coins actually available for payments and from how Coin Metrics defines adjusted transfer value and free float.
Transition risk matters more than static levels. The important shocks are vesting cliffs, emissions, staking withdrawals, collateral liquidations, bridge releases, and wrapper redemption events. Ethereum staking withdrawals became enabled on April 12, 2023, and rewards above 32 ETH are swept automatically every few days, so staked supply is not a permanent sink. It is a scheduled-release state with queue mechanics.
A stress-testing framework should therefore run in regimes. A base case assumes normal transfer behavior and scheduled unlocks. An event case layers in emissions campaigns, exchange listings, or governance changes. A stress case forces synchronized exits, wrapper discounts, and validator withdrawals. That structure is an inference, but it fits both the BIS exchange-rate equation and the fact that major supply releases in crypto usually arrive through rule-based schedules rather than smooth continuous flow.
Holding time should be modeled alongside velocity. Coin Metrics defines coin days destroyed as time multiplied by money, which makes it useful for separating old dormant balances finally moving from ordinary high-frequency churn. A token with unchanged headline velocity but rising age-destroyed supply is not stable. It may be preparing for a regime shift in available float.
Choose sinks that survive contact with users
Effective token sink mechanisms only work if they reduce economically spendable float without being trivially bypassed. Ethereum’s EIP-1559 permanently burns the base fee and reserves only the priority fee for validators, which reduces supply directly instead of relying on voluntary lockups alone.
Vote-escrow designs go further by combining time commitment, governance power, and value accrual. Curve’s veCRV is non-transferable, requires locking CRV, and allows a maximum lock of four years. Curve also distributes fees weekly to veCRV holders, and the original DAO design described reward boosts of up to 2.5x for users who vote-lock. Those mechanics clearly reduce float and strengthen holding incentives.
Every sink needs a leakage analysis. Curve’s original whitepaper explicitly notes that lock accounts should not be smart contracts because they can become tradable or tokenized. That is exactly the right modeling instinct. If a lock can be wrapped, borrowed against, or mirrored with a liquid receipt, legal lockup and economic velocity diverge.
Liquid staking is the clearest modern example of that divergence. Ethereum staking locks ETH at the consensus layer, but Lido mints rebasing ERC-20 stETH when users submit ETH and burns it when redeemed, restoring transferability and DeFi composability while the underlying ETH remains staked. Any model that counts all staked ETH as economically inert will understate real float.
Emissions should be modeled as a velocity source, not just a growth lever. Curve’s whitepaper states that first-year CRV inflow into circulating supply was approximately 2 million CRV per day and that emissions could be boosted via vote-locking. When recipients are yield farmers, market makers, or treasury actors with short holding horizons, a large share of new supply should be assumed to recycle rapidly into the sell side.
Regulatory constraints should shape the design
Regulatory treatment is not a side issue in velocity design. The SEC’s April 3, 2019 framework says a digital asset is more likely to create a reasonable expectation of profit when holders share in enterprise income or profits, receive dividends or distributions, or rely on secondary markets to realize gains.
That creates a real trade-off. Direct revenue sharing, protocol fee distributions, and explicit yield promises can materially slow turnover, but they also blur the line between a utility token and an instrument that markets a claim on enterprise cash flows. From a regulatory-pragmatist standpoint, this is design flexibility purchased with legal exposure.
Governance rights alone are not a compliance shield. The SEC framework focuses on economic reality rather than labels, and Curve’s own documentation shows how governance, fee distribution, and reward boosts can be bundled inside one token design. Once governance power also determines access to fees or emissions, the token starts to look less like a coordination primitive and more like a structured economic right.
The EU’s MiCA is not identical to U.S. securities analysis, but it pushes toward the same product clarity. MiCA defines a utility token as a crypto-asset only intended to provide access to a good or service supplied by its issuer. The regulation has applied broadly since December 30, 2024, with the asset-referenced and e-money token rules in force since June 30, 2024. If retention depends more on yield and fee participation than on service access, the utility narrative gets harder to defend.
The SEC framework also points to the opposite direction. It says a token is less likely to satisfy Howey when it is offered for use or consumption, when holders can immediately use it for intended functionality, and when the token’s value shows a direct and stable correlation to the value of the good or service for which it is exchanged or redeemed. That is a useful design constraint for teams that want lower velocity without leaning on quasi-equity rights.
As a design inference, fee burns and tightly scoped access utility are often easier to defend than explicit profit-sharing because they can create scarcity or usage demand without granting a direct right to distributions. Ethereum’s fee burn is the clearest live example of that structure. The trade-off is that the retention effect may be weaker, slower, or less legible to the market than a formal cash-flow claim.
A practical rule for token economy design
The operational rule for token economy design is simple: model velocity as a float-and-friction system, then choose the lowest-risk mechanism that can keep spendable float aligned with real transactional demand. The common error is not “too much velocity” in the abstract. It is misclassifying which balances are actually spendable, which incentives are immediately recyclable, and which retention tools create new jurisdictional problems.
Use adjusted economic transfer volume, not raw transfer counts or gross on-chain volume, as the numerator.
Use spendable float, not fully diluted supply and not naive circulating supply, as the denominator.
Separate hard locks from liquid wrappers. stETH-style receipts preserve economic liquidity even when base assets are staked.
Treat vesting schedules, emissions, and withdrawal mechanics as transition probabilities between supply states, not as static footnotes.
Prefer retention mechanisms tied to real utility first. Add governance sparingly. Add cash-flow rights only when the legal perimeter, disclosures, and jurisdiction strategy can carry them.
Track cohort behavior, because aggregate velocity can hide intermediary churn and concentration.
At FinDaS, our tokenomics consulting work usually treats velocity as a joint problem of market microstructure, token utility, and legal classification rather than a single dashboard ratio. That framing is not promotional. It is what the evidence supports. Tokens that require aggressive yield engineering to stop immediate resale often reveal a utility-demand problem first and a compliance problem second, and the ordering matters because the legal risk is frequently created by the mechanism chosen to patch weak retention.
