Fake trading volumeis what you get when the most quoted statistic in markets meets the lowest cost of manufacture. Volume drives exchange rankings, token screeners, liquidity assumptions, and market-size headlines, yet in unregulated venues it is self-reported, and on public blockchains it can be printed by one actor trading with themselves. This page is a working guide to consuming volume numbers without inheriting someone else's manipulation.
Where fake volume comes from
- Venue self-inflation. Exchanges printing their own tape to climb rankings, quantified in the academic record covered on crypto wash trading.
- Reward farming. Traders manufacturing volume because a token programme pays for it, the trade-to-earn pattern that consumed NFT markets in 2022.
- Tape painting. Classic wash trading to make an asset look alive ahead of a distribution.
- Sybil usage theatre. Fleets of wallets simulating adoption ahead of airdrops, which contaminates protocol usage metrics the same way.
The auditor's checklist
Before trusting any volume number, run it through five questions:
- Who reports it, and what do they gain? A venue ranked by its own self-reported figure has already answered your question.
- Does volume scale with funded participants? Divide volume by unique, independently funded actors. Manufactured markets show enormous volume per funder; organic ones do not.
- What survives self-trade filters? On-chain, remove round trips and buyer-funded-by-seller transactions and see what remains. The techniques are on the detection page.
- Do the statistics look organic? First-digit distributions, size roundness, and timing regularity separate human order flow from scripts printing a target number.
- Does price agree? Genuine volume moves and is moved by prices. Volume that towers over a flat, thin market is decoration.
What my data says about trusting the tape
The reassuring finding in the research: markets are better at ignoring fake volume than the manipulators hoped. Across NFT collections, wash volumes showed no significant relationship with genuine future volumes; buyers did not, in aggregate, chase the fake prints. The unflattering corollary: the fake volume kept being printed anyway, because venues were paying for it directly. Both facts belong in your model of any market statistic: the crowd is hard to fool for long, and the statistic will still be corrupted wherever corrupting it is subsidised.
For the NFT-specific numbers, see NFT wash trading. For the umbrella view of everything that distorts these markets, see NFT market manipulation.