Token velocity is how many times an average token changes hands over a period. It comes from the classical equation of exchange applied to crypto, where money supply times velocity equals transaction volume, and Vitalik Buterin's 2017 restatement of it puts market cap equal to daily economic value transacted multiplied by how long a user holds the token before spending it. High velocity means holders pass the token along quickly, which caps the market cap that a given level of activity can support.
Velocity is not a dial a designer turns. It is an output of how attractive holding is relative to spending, and the only thing a mechanism can change is that comparison. Any design memo that treats velocity as a direct control has skipped the argument.
Scroll to see the full diagram
The identity, and what it does and does not prove
Buterin's essay on medium-of-exchange token valuations sets out the standard equation: money supply multiplied by velocity equals price level multiplied by transaction volume, where velocity is the number of times an average coin changes hands each day and transaction volume is the economic value of transactions per day.1
He then recasts it into the form worth remembering. Substituting holding time for the inverse of velocity and token price for the inverse of the price level gives market cap equal to economic value transacted per day multiplied by the time a user holds a coin before transacting with it.1 Written that way, the market cap a payment-style token can sustain depends on two things: how much value flows through it, and how long anyone is willing to sit on it.
Note carefully what this is. An identity, not a causal model. It tells you the four quantities are related by definition. It does not tell you which one moves when you change a mechanism, and that gap is where most tokenomics arguments go wrong.
Where the fixed supply pitch breaks
Kyle Samani opened Multicoin's 2017 piece on this by quoting the pitch nearly every token deck of that era contained: there is a fixed supply of tokens, so as demand for the token increases, the price must increase. His response is that the logic fails to take into account the velocity problem.2
His worked example is event ticketing. Ticketing on a blockchain creates real value for venues, artists and buyers by removing fraud and cutting intermediary fees. But a buyer acquires the token, buys the ticket, and the venue converts out. Nobody in that chain has a reason to hold. Samani's argument, written in December 2017 about the ticketing tokens then in market, was that even platforms processing very large transaction volumes would not see the token price move materially, because the token mechanics gave nobody a reason to keep it.2
That is the shape of the problem for any pure medium-of-exchange design. The protocol can be useful, widely adopted and profitable for the operator while the token that runs through it captures almost nothing, because it is a toll booth rather than a store.
The rebuttal to read before you design around it
In December 2018 Brave Software commissioned Scott Locklin to examine the velocity argument directly. His report considers various assertions about the effect of token velocity on token exchange rate, and states in its abstract that the claim velocity is an extrinsic quantity which can be throttled to raise the exchange rate, or will naturally rise and lower it, is examined and rejected.3 A section of the report addresses Buterin's essay by name.
You do not have to agree with Locklin to benefit from reading him. The useful part is the discipline he applies: velocity appears in an identity, so treating it as an independent variable you can push on assumes exactly what needs to be proven. If a mechanism raises holding time, something else in the identity moves too, and a design memo that shows one term changing and holds the rest fixed is not an argument.
In our view this exchange is the most productive disagreement in tokenomics, and both sides are usually cited by people who have read neither.
What a sink actually changes
A staking lockup, a fee burn, a collateral requirement and a governance escrow are all described as velocity reduction mechanisms. None of them reduces velocity directly. What each one does is alter the payoff from holding relative to spending, and holding time responds to that comparison.
The distinction matters because it tells you when a sink will fail. A staking lockup funded by emissions raises holding time only while the emissions last, because the reason to hold is the emission rather than the protocol. Lock up a token to earn more of the same token and you have moved supply forward in time without creating any demand for it.
A sink holds when the reason to hold comes from something outside the token. Fees paid in a scarce asset and burned. Collateral required to do work the network genuinely needs. Access to something with independent value. The NBER model of token-based platform finance treats issuance and burning as levers tied to platform productivity rather than as free parameters, which is the right frame: supply mechanics are financing decisions, and they only work when there is production to finance.4
Where founders go wrong with this
The most common failure is starting from the mechanism. A team decides the token needs a sink, adds staking with an emissions-funded yield, and reports a high locked percentage as evidence the design works. The lock rate is real. The demand is a subsidy the treasury is paying for, and it stops when the treasury does.
The second failure is measuring velocity from onchain transfer counts and treating the result as meaningful. Exchange internal transfers, bridge operations, market maker rebalancing and wash activity all move tokens without anyone using them for anything, and none of that is separable from the outside without careful filtering.
The third is the one that matters most. Velocity questions get asked about tokens that had no reason to exist, at a point where the answer would need to be a different token design rather than a sink bolted onto this one.
Our position
We treat velocity as a diagnostic rather than a design target. If your token has high velocity, the interesting question is not how to slow it down but why nobody wants to hold it, and the answer is almost always that the business underneath does not route anything of value to holders.
That reframing changes the work. Instead of designing lockups, we look at what the protocol earns, who has a claim on it, and whether the token is the instrument through which that claim is expressed. When it is, holding time takes care of itself. When it is not, no sink will fix it, and adding one mostly delays the conversation.
The token is infrastructure for the business. Velocity is a symptom telling you whether the business is using it.
Common questions
What is token velocity?
Token velocity is how many times an average token changes hands over a period. Buterin's 2017 essay applies the classical equation of exchange to tokens, where money supply times velocity equals price level times transaction volume, then recasts it as market cap equal to daily economic value transacted multiplied by how long a holder keeps the token before spending it.1 High velocity means short holding times.
Why is high token velocity a problem?
Because a token that everyone passes along immediately captures very little of the value flowing through it. Kyle Samani's 2017 analysis used blockchain ticketing as the example: the platform can process enormous transaction volume while the token price does not move materially, because no participant in the flow has a reason to hold.2 The protocol works and the token does not benefit.
How do you reduce token velocity?
You cannot reduce it directly, which is the point of the main critique. A 2018 report commissioned by Brave Software examined and rejected the claim that velocity is an extrinsic quantity that can be throttled to raise a token's exchange rate.3 Mechanisms such as staking, burns and collateral requirements change the payoff from holding relative to spending. Holding time responds to that, and only while the reason to hold outlives the subsidy.
Is token velocity the same as trading volume?
No. Trading volume is the value exchanged on markets; velocity is turnover relative to supply or market cap. They also measure different things in practice, because onchain transfer counts include exchange internal movements, bridge operations and market maker rebalancing that are not anyone using the token for its purpose. Filtering those out is the hard part of measuring velocity at all.
See Tokenomics Design for how this applies in practice.
Sources
- On Medium-of-Exchange Token Valuations
Vitalik Buterin, 2017
Applies the equation of exchange to tokens and recasts it as market cap equal to daily economic value transacted multiplied by holding time. - Understanding Token Velocity
Multicoin Capital, Kyle Samani, 2017
Names the fixed-supply pitch, identifies the velocity problem, and works through blockchain ticketing as the case where volume does not reach the token. - Token Economics: Considering Token Velocity, Report BSI-02
Scott Locklin for Brave Software, 2018
Examines and rejects the claim that velocity is an extrinsic quantity that can be throttled to raise a token's exchange rate. Addresses Buterin's essay by name. - Token-Based Platform Finance, NBER Working Paper 27810
National Bureau of Economic Research, 2021
Models issuance and burning as endogenous levers tied to platform productivity rather than free design parameters.
Last reviewed 2026-08
Know the terms but not sure how they apply to your project? That is what an engagement is for. We design, document, and stress-test the whole token economy inside the Tokenomics Data Room.
100+ projects advised. Complete tokenomics in 4 to 6 weeks.