Wash trading in prediction markets describes a scheme (and process) where one trader, or several accounts working together, execute orders on both sides of the same trade. This mechanism generates public ledger entries without transferring economic risk or altering net market exposure for them.
Volume is one of the core metrics market participants analyze to verify activity levels before entering an order, and wash trades directly inflate reported volume figures while providing zero organic order flow or executable liquidity.
What is wash trading?
The Commodity Futures Trading Commission defines wash trading as: "Entering into, or purporting to enter into, transactions to give the appearance that purchases and sales have been made, without incurring market risk or changing the trader's market position."1CFTC, "CFTC Glossary: Wash Trading," cftc.gov, accessed September 2026
"The Commodity Exchange Act prohibits wash trading."2CFTC, "CFTC Glossary: Wash Trading," cftc.gov, accessed September 2026
In plain terms, the participating entity finishes with its original net position and capital balance, despite the exchange logging a completed transaction.
Individual venue rulebooks mirror federal law by explicitly banning self matching and non economic trading:
- Kalshi, Rule 5.17: "No Person shall engage in any trading activity intended to accomplish a 'money pass', 'wash trade' or 'front-running'".3KalshiEX LLC, "Rulebook, Version 1.29" (Rule 5.17), kalshi-public-docs.s3.amazonaws.com, 2026
- Polymarket US, Rule 7.2: Bans orders "intended to avoid taking a bona fide market position exposed to market risk," and states participants "must not intentionally match" their own orders.4Polymarket US, "Polymarket US Rulebook" (Rule 7.2), polymarketexchange.com, April 2026
- Polymarket Market Integrity Policy: Lists wash trading among prohibited practices alongside self dealing and fictitious transactions.5Polymarket, "Polymarket Policy: Market Integrity," integrity.polymarket.com, accessed September 2026
How much prediction market volume is wash trading?
Empirical studies of major prediction platforms show wash trading estimates ranging from a 1% median per market to nearly 25% of Polymarket's total historical volume:
- A Columbia University study reported by CoinDesk found nearly 25% of Polymarket's total historical volume came from users rapidly buying and selling contracts.6CoinDesk (Helene Braun), "Polymarket's Trading Volume May Be 25% Fake, Columbia Study Finds," coindesk.com, November 2025 Wash volume peaked at nearly 60% of weekly platform volume in December 2024, and in some weeks over 90% of trades in sports and election markets looked suspicious.7CoinDesk (Helene Braun), "Polymarket's Trading Volume May Be 25% Fake, Columbia Study Finds," coindesk.com, November 2025 Researchers flagged wallets that opened and quickly closed positions, mostly trading with other wallets that showed the same pattern.8CoinDesk (Helene Braun), "Polymarket's Trading Volume May Be 25% Fake, Columbia Study Finds," coindesk.com, November 2025
- A microstructure study of the Polymarket order book by Dubach measured a median self-counterparty wash share of 1% per market, with an upper tail of 22%, a considerable range below the 25% to 70% figures measured on unregulated crypto exchanges.9Dubach, "The Anatomy of a Decentralized Prediction Market: Microstructure Evidence from the Polymarket Order Book," arxiv.org, April 2026
These figures diverge because the respective research teams evaluated different time horizons and different detection methods.
Track events you care about. Set up free alerts in 30 seconds.
Why would anyone trade with themselves?
In financial markets, participants execute wash trades to project false trading activity, manipulate price discovery, or draw outside liquidity into illiquid contracts without holding net exposure.
In prediction markets specifically, traders also use self-matching trades to harvest volume-based incentives such as token airdrops. Generating high nominal turnover lets accounts qualify for these rewards while neutralizing their directional market risk.
Academic findings point to reward extraction as a likely driver of this activity on decentralized platforms. For example, a Columbia University study analyzing prediction market order flow found that suspected wash trading accounts "made no real profits" from market outcomes.10CoinDesk (Helene Braun), "Polymarket's Trading Volume May Be 25% Fake, Columbia Study Finds," coindesk.com, November 2025 Researchers suggested the goal "may have been to game future incentives like token airdrops," citing speculation surrounding a prospective Polymarket token distribution as a possible incentive.11CoinDesk (Helene Braun), "Polymarket's Trading Volume May Be 25% Fake, Columbia Study Finds," coindesk.com, November 2025
What does wash trading mean for an everyday trader?
Wash trading distorts exchange data by overstating how many independent buyers and sellers are participating in a given contract. Headline turnover can give a market the appearance of active trading while genuine liquidity remains minimal.
A wash trade does not provide an executable counterparty for other participants because the order opens and closes within the same account or coordinated group.
When wash trading masks a thin market, everyday traders face direct impacts:
- Higher trading costs: A thin order book lacks enough resting offers at the current price. When you place an order, it fills across worse price levels, forcing you to buy higher or sell lower than expected.
- Trouble exiting a trade: Entering a position during high volume looks easy, but leaving a trade requires a real buyer or seller. In a fake volume market, you can end up stuck in a contract or forced to take a loss to cash out.
- Sudden price swings: In a real market, ordinary trades barely move the price. In an empty market, even small orders cause big price jumps because there are few real offers to absorb the trade.
- Distorted probability signals: Prediction market prices reflect crowd probability. Wash trades can print prices no independent buyer and seller agreed on, briefly pushing the displayed price away from true consensus.