High-frequency trading is a method of executing large numbers of orders at very low latency. It is not inherÂently market manipÂuÂlation. The invesÂtigative question is whether a strategy was designed to create a false impression of supply, demand or price, or whether rapid order changes reflected legitÂimate market making, arbitrage or risk control.
Define the suspected conduct precisely
Start with the instrument, venue, particÂipant, account, algorithm and time window. State the alleged mechanism: spoofing or layering, marking the close, wash trading, momentum ignition, cross-market manipÂuÂlation or misuse of confiÂdential order inforÂmation. Do not use a high cancelÂlation rate or fast execution alone as proof of intent.
Preserve complete order, modifiÂcation, cancelÂlation and execution messages with synchroÂnised timestamps. Add market data, auction imbalÂances, positions, profit and loss, risk limits, algorithm versions, deployment records and trader commuÂniÂcaÂtions. Our guide to analysing transÂaction patterns explains the broader principle: a reproÂducible event sequence is stronger than a list of anomalies.
Reconstruct the order-book sequence
For each suspected episode, show what the particÂipant could see, which orders it placed, where they sat in the queue, whether they were executable, how long they remained, what traded on the opposite side and when the larger orders were cancelled. Compare this with the particÂiÂpant’s normal behaviour and with other firms under the same volatility and liquidity condiÂtions.
Spoofing generally involves orders placed with an intent to cancel before execution so that other particÂiÂpants receive a misleading signal. The SEC’s litigation against Lek Securities and related defenÂdants described layering or spoofing as placing and cancelling orders to induce trades at artificial prices. That finding followed a trial; it should not be converted into a shortcut for judging unrelated order data.
Test intent and economic purpose
Examine fill rates, distance from the best price, order lifetime, size, repetition, cancelÂlation timing and direcÂtional changes. Then test legitÂimate explaÂnaÂtions: inventory management, stale-price protection, venue fragmenÂtation, a news event or a change in displayed liquidity. Internal code names, design documents, messages and consistent profitability on the opposite side can strengthen an intent analysis.
In its first HFT manipÂuÂlation case, the SEC found that Athena Capital Research manipÂuÂlated closing prices through aggressive last-second trades and cited contemÂpoÂraÂneous internal messages. The important lesson is evidential: the case combined market data, strategy design, price impact and commuÂniÂcaÂtions rather than treating speed itself as misconduct.
Measure impact without exaggeration
Estimate artificial price movement, affected volume, execution harm and profit using a documented counterÂfactual. Control for news, index rebalÂancing, auctions and wider market moves. Replicate the analysis across compaÂrable days and have an independent quantiÂtative reviewer test the code and assumpÂtions.
Michael Schmitt’s discussion of market-manipÂuÂlation indicators offers useful practiÂtioner context on order-book anomalies. InvesÂtiÂgators should still base legal concluÂsions on the applicable market-abuse rules, primary trading records and regulator guidance.
Build a defensible case file
Create an episode table linking every allegation to raw messages, derived metrics, commuÂniÂcaÂtions, alterÂnative explaÂnaÂtions and review notes. Preserve exculÂpatory evidence and disclose model limitaÂtions. Separate surveilÂlance alerts, which prioritise review, from findings capable of supporting enforcement.
The correct conclusion may be manipÂuÂlative conduct, a control weakness, poor algorithm design or legitÂimate trading. High-frequency trading becomes unlawful when the evidence estabÂlishes the prohibited conduct and required intent under the relevant regime—not merely because the activity was fast, complex or profitable.