Wash trading is the practice of buying and selling the same asset with yourself, either directly or through accounts under common control, so that the market records activity that involves no genuine change of ownership. The tape shows trades, volume, sometimes a rising price. Economically, nothing happened. The point is the appearance: to third parties the asset looks liquid, traded, and wanted.
The technique is much older than crypto. US law has prohibited wash sales in commodity futures since the Commodity Exchange Act of 1936, and bucket-shop era stock pools used matched orders to paint the tape decades before that. What changed with crypto and NFT markets is the cost structure: creating a second identity went from opening a brokerage account to generating a wallet address, which is free and takes seconds.
The mechanics
A wash trade needs three ingredients, each trivial in unregulated markets:
- Two trading identities under one decision-maker. On an exchange, two accounts; on-chain, two wallets funded from the same source.
- A venue that reports activity. Volume tickers, leaderboards, trending pages, or any statistic that other people use as a demand signal.
- A payoff that scales with reported activity. Buyers lured by the fake signal, rewards paid per unit of volume, or a cleaner-looking price history.
In order-book markets the classic form is the matched order: one account posts a bid, the sibling account hits it, and the pair repeats. In NFT markets the form is even simpler, because trades are bilateral: wallet A sells the token to wallet B, B sells it back or on to wallet C, and all three wallets trace back to one funding source. The detection page covers how those traces are found in practice.
Why traders do it
In The Economics of Wash Trading I split the motives into two testable families, and the distinction turns out to matter more than the headline volumes:
- Attractiveness. The manipulator wants real buyers. Fake volume and managed prices exist to make a collection or coin look like the crowd is already there. This is the folk theory of wash trading, and it is the one the data supports least.
- Token-based incentives. The manipulator wants the reward attached to trading itself. When a marketplace pays its own token per dollar of volume, every wash trade is a small claim on the emission schedule. The trade is profitable even if no third party ever looks at the chart. The LooksRare episode is the cleanest natural experiment here.
My central finding: in NFT markets, wash trading volumes showed no significant relationship with real trading volumes on future days, while wash activity concentrated exactly where token incentives paid for it. Most wash trading, in other words, is not failed advertising. It is successful farming.
How big is it?
Estimates vary by market and method, but the consistent result is that unregulated venues carry a large multiple of their genuine activity. Cong, Li, Tang and Yang's Crypto Wash Trading estimated that well over half of reported volume on many unregulated crypto exchanges was wash traded. For NFTs, von Wachter and co-authors flagged a small share of transactions but a large share of value on specific days, and during reward programmes the fake share of volume became the majority. I collect the NFT-specific numbers on the NFT wash trading page.
Wash trading vs the tax wash sale
Two different things share the word wash. The wash sale rule is a tax provision: under IRS rules, selling a security at a loss and repurchasing a substantially identical one within 30 days disallows the loss deduction. It involves real trades with the market and is legal, merely tax-ineffective. Wash trading is trading with yourself to manufacture activity, and in regulated markets it is a prohibited manipulation. Whether the prohibition reaches crypto and NFTs is its own question, answered at length on Is wash trading illegal?
Why it persists
Wash trading survives because the signal it corrupts, volume, is the cheapest signal markets have. Rankings sort by it, aggregators chart it, incentive programmes pay for it, and newcomers read it as social proof. As long as anything of value keys off reported volume, someone will manufacture reported volume. The durable fixes are structural: reward mechanisms that pay for something harder to fake, statistics that exclude self-financed flow, and market designs covered on the NFT markets page. The enforcement picture is improving too, but slowly, and mostly after the fact.