Collusion set detection using graph clustering

Abstract

Many mal-practices in stock market trading-e.g., circular trading and price manipulation-use the modus operandi of collusion. Informally, a set of traders is a candidate collusion set when they have "heavy trading" among themselves, as compared to their trading with others. We formalize the problem of detection of collusion sets, if any, in the given trading database. We show that naïve approaches are inefficient for real-life situations. We adapt and apply two well-known graph clustering algorithms for this problem. We also propose a new graph clustering algorithm, specifically tailored for detecting collusion sets. A novel feature of our approach is the use of Dempster-Schafer theory of evidence to combine the candidate collusion sets detected by individual algorithms. Treating individual experiments as evidence, this approach allows us to quantify the confidence (or belief) in the candidate collusion sets. We present detailed simulation experiments to demonstrate effectiveness of the proposed algorithms.

Detecting price manipulation in the financial market

“…Due to a few real manipulation cases being reported, we needed to synthesize a number of abnormal cases based on our study of the characteristics of the manipulation strategy. Synthetic exploratory financial data is accepted in academia for evaluating the proposed model when real market data is hard to collect [30] [31].…”

On Wash Trade Detection in Energy Markets

“…Authors assume that there is a strong correlation between trader's current behaviors and his/her previous trading network. A graph clustering algorithm for detecting a set of collusive traders has been proposed in [7]. Some authors believe it is unacceptable to ignore the order price information, which not only distinguishes traders' intention, but is a key feature of wash trade activity [8], [9], [10].…”

Detecting wash trade in the financial market

“…The assumption of this work is the strong consistence of a trader's current behaviours and his/her previous trading network. A graph clustering algorithm for detecting a set of collusive traders has been proposed [5]. The relationship between traders is constructed as a stock flow graph, and those with "heavy trading" within their network are clustered as collusion set.…”

“…To date, academic research has mainly focused on the detection of trading collusions according to analogous trading behaviours [5] [6], which were defined by aggregated order sequences across various stocks. Detecting the overall behaviours across different stocks can hardly reach a This project is supported by the companies and organizations involved in the Northern Ireland Capital Markets Engineering Research Initiative.…”

“…Wash trades can be carried out by either one single trader or multiple traders [5] [6] [7]. The arrangement of one or various traders in a trading collusion is essentially analogous to a logically topological structure of various nodes in a network [9].…”