By Padraig Kelleher
KX has a broad list of products and solutions built on the time-series database platform kdb+ that capitalize on its high-performance capabilities when analysing very large databases.
KX for Surveillance is a robust platform widely used by financial institutions for monitoring trades for regulatory compliance. The surveillance platform instantly detects known trading violations like layering, spoofing or marking the close. Customers can calibrate their parameters in real time to improve their detection quality and accuracy. The flexibility of the historical database and replay engine eases retrospective investigation for new types of fraudulent and suspicious activity.
In this series, we take a look at what makes KX for Surveillance such a powerful tool for detecting market manipulation. In this article we will take a close look at Ping Order alerts.
Ping orders are a form of market manipulation in which traders enter small orders for the purpose of price discovery. It is a form of market abuse whereby a trading order rests on a dark platform in search of large hidden orders. They are typically used to locate and trade ahead of large buyers and sellers.
Ping orders try to exploit instances where algorithmic trading systems might break large orders into a series of smaller ones and feed them steadily into the market in order to reduce their cumulative impact. To that end small-sized “immediate-or-cancel” orders for a security may be sent to a trading venue to try to discover if there is demand for that security (either buy or sell) that is not otherwise visible. When a trader gets a “ping” (i.e. the trader’s small order has been executed) or series of pings that alerts the trader to the presence of a large buy-side order, it may engage in a predatory trading activity that ensures a nearly risk-free profit at the expense of the buy-side, who will end up paying an unfavourable price for its large order.
Authorities and exchanges such as SGX, SSE and SZSE have recognized ping orders as an area of market manipulation and have introduced regulation to protect markets and investors.
KX for Surveillance provides a robust solution for identifying ping orders. The kdb+ database that underlies it facilitates nanosecond (if available data allows it) analysis of every individual event, private and market, and will generate a ping order alert based on configurable rule sets. Sample rules might include:
- If a buy or sell cancel order was entered in configurable time window
- If the buy or sell ping order count exceeds a configurable threshold
- If the ping order price percentage exceeds a configurable percentage threshold
The screenshot above illustrates a Ping alert in operation. It can be seen that the trader has entered a number of small orders to ascertain the market and then shortly afterwards cancels the sell order based on market reaction. The alert was subsequently fired when the cancel time and percentage gain thresholds were both exceeded. The tracing volumes are illustrated in the bottom panel.
The value of what constitutes a large or small order can either be set to be a set amount or calculated via benchmarking on a per entity basis. Benchmarking is the use of historical data to obtain a tailored threshold value for each entity value based on their normal trading characteristics.
The surveillance pattern looks for multiple instances of this activity throughout the trading day in order to raise an alert. One reason for this is that the volumes traded during one instance can be quite low and therefore the potential profit is not significant. In contrast, multiple occurrences would result in higher potential profits, along with stronger evidence that the entity is deliberately attempting to manipulate the market utilising the practice of ping orders. The determination will ultimately be made by alerts analysts as to whether or not the unusual activity was malicious and if there was an attempt made to manipulate the market.
For more information on KX for Surveillance and its functionality please click on the links below.
