Abusive Messaging Detection

Elise Fleischaker Product Features

scorefeature
The
Abusive Messaging Detection model identifies patterns of disruptive or excessive order activity.

Quote Stuffing

Quote stuffing is a messaging pattern designed to disrupt normal market operations by introducing latency into an exchange’s quoting engine. This latency prevents other market participants from trading as intended, and can be used as a component of a number of abusive messaging strategies.  The Neurensic Abusive Messaging Detection identifies quote stuffing using a dynamic pattern detection algorithm that eschews the traditional measurements of messages per second in favor of intensity and potential market harm. This novel approach detects intense bursting of order message regardless of duration, addressing the fact that an even relatively small number of orders can introduce latency into the markets if compressed into a short enough period.

Microburst Quote Stuffing

The traditional approaches to detecting abusive messaging interrogate messaging rates over arbitrary intervals of time, most commonly one second. This method is unable to detect extremely high rates of messaging over very short periods of time. These microbursts of activity can be used to introduce predictable latency into the market while avoiding traditional detection methods. For instance, an intense burst of 80 messages in a single millisecond will appear like a relatively sluggish 80 messages per second–despite having an effective rate of 80,000 messages per second. The higher the messaging rate, the more likely it is that a predictable latency could be introduced into the market.

The model measures messaging rates, looking for high rates regardless of duration, and allows users to evaluate the cause/effect relationship between the incident and possible harm.

Malfunctioning Algos

Malfunctioning algorithms have been responsible for numerous disruptive market events, and regulators have committed significant resources to detecting and bringing actions against firms that are responsible for them.Similar to the method for detecting quote stuffing, regulators identify incidents where messaging rates are above a certain order per second threshold.

The Neurensic Abusive Messaging model goes a step further by evaluating potential problematic activity, regardless of the duration of the event, and presents the results such that users can judge if the pattern is a possible malfunction or makes sense in context.

In an industry where market participants battle over microsecond advantages, measuring abusive messaging using broad strokes leaves the market and compliance departments vulnerable. The Neurensic approach addresses this vulnerability and further safeguards the integrity of the markets–no matter the speed.

 

See Also: Quote Stuffing Disrupting Markets