Market Anomaly: Wash Trading

• Coinbase (GDAX), 2015–2018: The U.S. derivatives authority found that proprietary self-trade configurations occurred on GDAX; misleading volume reports were sanctioned. Result: Settlement of USD 6.5 million on March 19, 2021.
• Binance.US, 2019–2022: In the SEC complaint of June 5, 2023, it is alleged that an affiliated trading firm (Sigma Chain) used wash trades to inflate volume on Binance.US. Status: Regulatory allegation, case pending.

Briefly put: In wash trading, buy and sell orders from the same economic entity are executed against each other so that no real ownership change occurs - the “trading volume” looks large but is fake activity.

How it works in practice

1. One actor controls multiple exchange accounts (e.g. main and sub-accounts).
2. They place simultaneous buy and sell orders of the same size at the same price level.
3. The matching engine sees two matching orders and fills them - without knowing both sides belong to the same entity (if no effective Self-Trade-Prevention is active).
4. Repeated every few seconds or via bots, this creates the illusion of liquidity and demand while no new participants are involved.

Why it works: Many traders use volume as a trust signal (breakouts, listings, rankings). Fake volume can distort price discovery, attract algos, and temporarily affect spreads and depth. Prohibited in regulated markets; on CEX, prevention depends on internal surveillance (see STP).

• Self-trade patterns: Unusual number of fills between linked accounts (KYC/IP/key clusters), net position ≈ 0.
• “Ping-pong” sequences: identical sizes/prices in tight timing between the same account pairs.
• Volume without price progress: heavy tape activity while VWAP/spread barely move.
• Statistical anomalies: Clustering of round sizes and Benford-law deviations - robust indicators in research.

CEXs control their own order books and reporting; external verifiability of order origins is limited. There are also conflicts of interest (e.g. volume rankings, market-maker programs). The NY Attorney General documented transparency and supervision gaps on crypto platforms in 2018.

On stock markets, Self-Trade Prevention (STP/SMP) is standard (e.g. Nasdaq) and automatically prevents orders from the same participant matching against each other. Additionally, the Consolidated Audit Trail (CAT, SEC Rule 613) enables complete order tracking across venues - circular self-trades are spotted faster and sanctioned.

Early detection protects traders from fake validity in volume and misjudgment of liquidity. In the EU, MiCA has been applicable since December 30 2024; for providers already active before, a transitional period runs until July 1 2026 (depending on the member state). ESMA Guidelines 2025 specify how market abuse – including volume-based deception – is to be detected and monitored preventively, imposing stricter requirements on surveillance systems and documentation.

Adaptive, quantile-based, measured per symbol / venue against a 30-day baseline at the same time of day. Required data: trade-tape time / price / size. Ideally: venue flag for self-trades.

E3 (high) – strong evidence

• Round-size ratio (share of round sizes like 1 / 2 / 5 / 10 / 20 × lot) > p99 and
• Benford deviation (first-digit distance) > p99 and
• VWAP drift within 15-min window < 5 bp with volume > p99.

E2 (medium)

• Serial “A↔B” sequences (recurrent cross-trades between two clusters) > p99 or
• Self-trade ratio (if available) > p99 with net position change ≈ 0.

E1 (low)

• Round-size ratio > p97 or Benford deviation > p97 (without volume extreme) → Watchlist.

The statistical logic – rounding / Benford – is well-established in academic literature on crypto wash trading.

• Differentiate from market-making: Internal netting or inventory shifts can simulate ping-pong; evaluate account clusters (KYC / IP / API keys).
• Check news / index events: Re-weights, airdrops, fee or tick changes can temporarily cause atypical size patterns.
• Cross-venue comparison: Examine price path / volume patterns across venues; avoid pure venue artifacts.

• For traders: Avoid fills on inflated volume, reduce slippage risk, assess trend quality more realistically.
• For operators / compliance: Well-documented signals (metrics, timestamps, baselines) facilitate internal reviews, customer communication and, if needed, cooperation with regulators.

• U.S. Commodity Futures Trading Commission (CFTC) – Press release “CFTC Orders Coinbase Inc. to Pay $6.5 Million for … Wash Trading”, March 19 2021 (officially determined / settled).
• CFTC – Order “Coinbase, Inc.”, March 19 2021 (detail on internal programs “Hedger” / “Replicator” and self-trades on GDAX).
• U.S. Securities and Exchange Commission (SEC) – Complaint vs. Binance / BAM Trading / Sigma Chain, June 5 2023 (regulatory allegations; section on “Sigma Chain’s Wash Trading”).
• Office of the New York State Attorney General (NY OAG) – “Virtual Markets Integrity Initiative Report”, September 18 2018 (transparency / oversight deficiencies on crypto platforms).
• Nasdaq – “Self Trade Prevention (STP) – Factsheet”, April 2 2024 (operation of STP / SMP in regulated markets).
• SEC / CAT NMS LLC – Overview “Consolidated Audit Trail (CAT) – Rule 613” (cross-venue order audit trail in the U.S.).
• Cong, Li, Tang, Yang – “Crypto Wash Trading”, NBER Working Paper 30783 (2022); Management Science 69 (11), 2023 (Benford / size tests; evidence on 29 CEXs).
• ESMA – MiCA site & 2025 guidelines on market-abuse prevention / detection (transition rules, EU supervisory practice).