How Regulators Spot Fake Volume on Crypto and FX Venues

Markets run on a simple assumption: when a price moves, someone somewhere genuinely wanted to trade. Volume is the market’s body language. When it looks busy, people lean in. When it looks quiet, they wait.

That is exactly why some participants try to fake the body language. They create trades that look real but are not, or orders that look like interest but are never meant to be filled. Three of the best-known versions are wash trading, self-matching and spoofing. Once you can see how each works, and how exchanges and regulators spot them, the whole idea of “artificial volume” becomes much less mysterious.

The shared trick: making activity look bigger than it is

Wash trading is the umbrella term. In plain words, someone buys and sells the same instrument in a way that creates the appearance of trading without any real change in who owns what. In the UK, the Market Abuse Regulation describes it as a sale or purchase where there is no change in beneficial interest or market risk, or where that risk only moves between parties acting in concert or collusion, other than for legitimate reasons.

The goal is rarely about price. As surveillance specialists point out, the aim is to raise the volume traded for an instrument so it appears more liquid, drawing other people in. The trades are typically very close in price and quantity and have no real economic purpose. This family of conduct goes by many names: wash sales, circular trading, pre-arranged trading, parking, churning.

Self-matching is the version where you are effectively on both sides yourself, or you and a cooperating counterparty line up a trade in advance. In a UK enforcement case, a senior market-making trader monitored how many shares of a real-estate company were legitimately traded each day. When the count fell below what he believed was needed to stay in the FTSE All Share Index, he executed buy and sell orders simultaneously to fabricate trades. He was fined £75,000 and banned from working in the securities industry.

Spoofing is different. It is not about completing a fake trade; it is about the order that was never meant to happen. The CFTC defines it as bidding or offering with the intent to cancel the bid or offer before execution. Think of it as putting a large order in the queue, letting the market react to the apparent depth, then pulling it. The intent to cancel is what separates spoofing from a genuine change of mind. A single instance can be enough.

What surveillance actually looks at

Regulators and exchanges do not read minds, so they look for patterns that reveal intent. Several checks come up again and again.

The first is the “no change in position” test. If a trader buys and sells the same quantity within a very short window at very similar prices, and their position and profit-and-loss barely move, that is worth a question. Normal trading changes something: your exposure, your inventory, your hedge. A flurry of trades that leaves everything exactly where it started is a classic wash-trading signature.

The second is order placement around specific windows. In the US futures case, traders entered bids and offers during the pre-open period, then tried to cancel them during the “No-Cancel” or Lockdown period, which is typically the last 30 seconds before the market opens. Rules blocked the cancellation. So they entered offsetting orders during that same lockdown, and those offsetting orders cancelled out the originals immediately after the open. The firm paid a $350,000 civil penalty, and the exchange added a $150,000 fine.

The third is the relationship between price and volume. Genuine markets usually show a rational relationship between consecutive prices and a reasonable correlation between price changes and volume. Spoofing often shows the opposite: big apparent depth that vanishes the moment real liquidity arrives, or price jumps that do not match the volume behind them.

The fourth is data completeness. Surveillance is only as good as its inputs. Regulators have repeatedly flagged that firms miss cancelled and amended orders, and those orders are often the whole point, since conduct that never results in a trade can still send false signals. Under-detailed alert settings are another recurring weakness: apply the same threshold to a blue-chip stock and a thinly traded bond, and you will either drown in noise or miss the abuse.

And increasingly, machines do the watching. The UK regulator receives millions of transaction and order reports each day, and that volume is analysed with data analytics. In crypto, deep liquid markets are dominated by electronic surveillance of order books and wallet-level patterns. Competing layers of surveillance exist for good reason.

The honest but confusing part: hedging and rolling look similar

Here is where careful reading matters. Many lawful trades look, on the surface, like the behaviour above. A dairy processor hedging a future purchase will enter and later offset futures positions. A market maker adjusting inventory will buy and sell the same contract repeatedly. Firms roll positions from one delivery month to another, which the CFTC glossary calls a spread.

What separates hedging from manipulation is not the shape of the trade. It is everything around it: whether the offsetting position manages a real business risk, whether it is documented, whether it is executed openly at competitive prices, and whether the participant is willing to have it explained. Wash sales and noncompetitive transactions are the opposite of that, precisely because they negate market risk and price competition instead of managing it. A hedge that genuinely manages exposure still leaves you with a position that behaves differently when prices move. A wash trade leaves you exactly where you began.

The legitimacy also depends on venue and jurisdiction. In the US, CFTC rules prohibit brokers from profiting from wash trading even if they say they were unaware of the trader’s intentions. In the UK, the FCA’s expectations for surveillance apply to regulated markets and investment firms, with spot commodities in scope only in certain situations. Crypto sits in the gap: some venues are regulated, some are offshore, and enforcement co-operation across borders is improving but uneven. Where a venue is registered, which regulator has authority, and which rules apply are genuinely important questions, and the answers are not the same everywhere.

Surveillance systems will always generate false positives. A legitimate algorithm that fires thousands of small orders a day may trigger volume alerts simply because it is busy. That is why regulators emphasize calibration, context and human review rather than auto-flagging. It is also why market abuse risk assessments are supposed to map specific risks to specific controls. One size fits all has never worked well. The same is true of many things in life, including chai.

The practical takeaway for anyone reading market activity, whether you trade crypto, FX or futures, is to treat volume as a signal that needs context. Unusually high volume with no news, no change in price direction and a suspiciously round number of trades deserves scepticism. So does apparent depth that vanishes the instant you try to trade against it.

Markets are not perfect at revealing who is real and who is not. But with order-book data, cross-venue comparison, improved analytics and enforcement action across the US, UK and elsewhere, artificial volume is harder to hide than it used to be. Understanding what regulators look for helps you read the signals more honestly, and helps you spot when something in the numbers just does not add up.