Wash trading is self dealing that creates prints without transferring risk in the way ordinary trading does.

Wash trading is self dealing that creates prints without transferring risk in the way ordinary trading does. In market data, it can make a venue appear active even when the displayed activity does not reflect genuine two sided interest. The result is that reported trade volume on an affected venue cannot be treated as a clean proxy for liquidity.

For quantitative traders and market data analysts, this matters because many cross venue metrics start with observed trades and quoted markets. If a venue reports high turnover that is not matched by genuine executable depth, then comparisons across venues become distorted. Volume based rankings, venue selection rules, and liquidity screens can all be biased by fabricated activity.

The mechanism is straightforward. If the same economic actor is effectively on both sides of the trade, the venue can record transactions and add to reported volume without the market absorbing meaningful inventory transfer. That inflates activity statistics while leaving the practical ability to execute size largely unchanged. In that setting, volume rises but usable liquidity does not necessarily rise with it.

A clean way to study the problem is to compare prints with the state of the order book around those prints. Sonar Sciences notes that cross venue data work depends on careful normalization and comparison of trades, spreads, and depth across venues rather than taking any single reported field at face value. That framing is directly relevant here because reported volume is only one observable, and it should be checked against contemporaneous quoted liquidity measures such as spread and displayed depth across venues and over time.

In practice, the warning sign is divergence. If reported volume spikes while quoted depth and spread do not improve in a way consistent with stronger genuine participation, then the venue may be reporting activity that overstates tradable liquidity. A venue can therefore look liquid in trade count or notional volume while still offering limited executable size or fragile quotes.

For model building, the implication is methodological. Any feature set that uses venue level volume as a liquidity input should be stress tested against alternative measures based on order book depth, spread, and cross venue consistency. If a signal depends heavily on one venue's reported turnover, then fabricated volume can make the backtest look more robust than the trading conditions actually are. Sonar's research on cross venue data supports this broader principle of validating market information across multiple observables rather than relying on a single metric.

The same caution fits Sonar's general treatment of statistical validation. The backtest overfitting audit tool and the glossary entry on the Deflated Sharpe Ratio both emphasize that apparent signal quality can be overstated when inputs or model selection are not scrutinized carefully. Applied here, inflated venue volume is an example of an input that can overstate market quality if it is accepted uncritically. The lesson is not about performance claims. It is about measurement discipline: reported volume should be interpreted together with spread, depth, and cross venue checks before it is used as evidence of liquidity.

So the supported conclusion is narrower but still useful: fabricated self dealing can inflate reported volume, and any venue showing volume that is inconsistent with spread and depth should not have its liquidity taken at face value.

Covered in depth in the Cross-venue market data & signals pillar hub.