Onchain velocity of money: does MV=PQ work for token economies?

Chris Burniske introduced the MV=PQ framework to crypto valuation in 2017. The idea was elegant: borrow the equation of exchange from monetary economics, plug in onchain data, and derive a fundamental value for any token. Eight years later, the framework remains one of the most debated valuation tools in crypto. It works sometimes. It fails spectacularly in other cases. Understanding why requires examining what velocity actually measures onchain and where the traditional equation breaks down.

The equation of exchange

Irving Fisher formalized the equation MV=PQ in 1911. M represents the money supply. V is the velocity of money, the average number of times each unit of currency changes hands in a given period. P is the price level. Q is the quantity of real economic output.

Rearranged for price: P = MV/Q. If velocity and output stay constant, increasing the money supply raises prices. If the money supply and output stay constant, increasing velocity raises prices.

Applied to tokens, the variables map differently. M becomes the circulating supply of the token. V becomes the onchain transaction velocity. P becomes the token price. Q becomes the total economic value the network facilitates. The rearranged form, P = Q/(MV), suggests that for a given level of economic output, higher velocity depresses price.

Measuring velocity onchain

Traditional economists estimate money velocity indirectly. The Federal Reserve calculates M2 velocity by dividing nominal GDP by M2 money supply. It's a rough aggregate. You cannot observe every dollar transaction in the economy.

Crypto makes velocity directly observable. Every token transfer exists on a public ledger. Velocity can be calculated as: total transfer volume divided by circulating supply over a given period. If 100 million tokens circulate and 200 million tokens worth of transfers occur in a quarter, quarterly velocity equals 2.

This transparency reveals dramatic differences across assets. Bitcoin's annualized velocity hovers between 4 and 7, meaning the average BTC changes hands roughly 5 times per year. Much of the supply sits in wallets untouched for years. Ethereum's velocity runs higher, between 8 and 15, driven by DeFi interactions, token swaps, and bridge activity. Stablecoins like USDT operate at velocities above 50, functioning as pure transaction media with minimal holding demand.

These measurements come with caveats. Internal exchange transfers, wash trading, and smart contract interactions inflate raw velocity figures. Adjusted velocity metrics that filter exchange-to-exchange transfers and contract calls produce lower, more meaningful numbers. The directional differences across asset classes remain consistent regardless of methodology.

The token velocity problem

Kyle Samani articulated the "token velocity problem" in 2017: utility tokens that users acquire only to immediately spend face relentless sell pressure. Nobody holds a subway token for investment purposes. You buy it, ride, and the recipient (the subway system) sells it back to the next rider. The token facilitates a transaction but generates no holding demand.

Applied to crypto utility tokens, this dynamic plays out consistently. Tokens used purely as gas or access credentials tend to trade at lower multiples of their network activity than tokens with built-in holding incentives. If a user needs 10 tokens to access a service, they buy 10, spend 10, and the service provider sells 10 on the open market. The token's velocity approaches infinity in the limit, and MV=PQ implies its price approaches zero relative to network output.

This isn't theoretical. Multiple utility token projects from 2017 to 2019 facilitated meaningful transaction volumes while their tokens collapsed in value. The network had users. The token didn't have holders.

Velocity sinks

If high velocity suppresses token price, the rational response is to design mechanisms that reduce velocity. The crypto ecosystem has converged on several approaches that function as velocity sinks.

Staking locks tokens for a defined period, removing them from circulating supply. A token with 70% of supply staked has an effective circulating supply of 30%, dramatically reducing the denominator in velocity calculations.

Governance rights create holding incentives beyond pure economic utility. If holding a token grants voting power over protocol upgrades, fee distribution, or treasury allocation, rational actors accumulate rather than transact. Curve's vote-escrow model (veCRV) takes this further: locking CRV for up to four years grants amplified voting power and boosted yield. The time-weighted lockup creates a direct, mechanical velocity reduction.

Burn mechanisms permanently remove tokens from supply. While this technically reduces M rather than V, the economic effect is similar: fewer tokens available for each unit of network output pushes the P = Q/(MV) equation toward higher prices.

Fee distribution turns tokens into productive assets. If holding a token entitles you to a share of protocol revenue, the token behaves less like a medium of exchange and more like an equity instrument. Holders earn real yield by not transacting, creating a natural brake on velocity.

Where MV=PQ breaks down

The equation of exchange was designed for national economies with a single currency and relatively stable institutional structures. Applying it to token economies introduces several fractures.

Tokens serve multiple functions simultaneously. ETH operates as gas (high velocity), a staking asset (near-zero velocity), collateral in DeFi (moderate velocity), and a speculative investment (variable velocity). Each use case implies a different velocity, yet all share the same supply. Aggregating these into a single V produces a number that doesn't accurately represent any individual use case.

Cross-chain activity complicates measurement. When ETH is bridged to Arbitrum and used in transactions there, does that activity count toward Ethereum's Q? The same token generates economic activity across multiple networks, and double-counting or undercounting is unavoidable with current measurement approaches.

Speculative demand dominates most token markets. A speculator buying ETH with no intention of using the Ethereum network contributes zero to Q but reduces V (the token sits in a wallet). Speculative holding demand can sustain prices far above what network utility alone would justify, at least temporarily.

Layer 2 scaling changes the relationship. As activity migrates to rollups, the base layer token may see lower direct velocity while the total economic output attributable to the ecosystem grows. This decoupling renders naive velocity calculations misleading.

What velocity data tells investors

Despite its limitations, velocity analysis remains useful when applied with precision rather than as a universal valuation model.

Relative velocity comparisons within the same asset class reveal meaningful patterns. If two competing DeFi governance tokens facilitate similar economic activity but one has 3x the velocity, the higher-velocity token likely has weaker holding incentives and may face more sustained sell pressure.

Velocity trend changes signal shifts in holder behavior. A declining velocity for a specific token often correlates with accumulation phases, as long-term holders absorb supply from short-term traders. Rising velocity during price declines suggests distribution.

The ratio of exchange-adjusted velocity to total velocity indicates how much trading activity is speculative versus utility-driven. A token where 80% of volume occurs on exchanges with minimal onchain utility transfers may be pricing speculation rather than fundamental demand.

For protocol designers, velocity analysis identifies whether a token model generates organic holding demand or merely forces users through an unnecessary intermediary step. If removing the token from the protocol's workflow would not meaningfully change user experience, the token likely has a velocity problem that no amount of engineering can fix.

Related Concepts

• Monetary policy of Layer 1s: How supply schedules interact with velocity
• Emissions vs revenue: Separating supply expansion from economic output
• DeFi liquidity trap: When velocity sinks fail and sell pressure dominates
• Real users vs subsidized activity: Distinguishing genuine velocity from incentivized transactions
• Bribes, rebates, and kickbacks: Vote-escrow mechanics as velocity sinks
• Bitcoin: Low-velocity store-of-value token dynamics
• Ethereum: Multi-function token with varied velocity profiles