Detecting Wash Trade in Financial Market Using Digraphs and Dynamic Programming (original) (raw)
Related papers
A wash trade refers to the illegal activities of traders who utilize carefully designed limit orders to manually increase the trading volumes for creating a false impression of an active market. As one of the primary formats of market abuse, a wash trade can be extremely damaging to the proper functioning and integrity of capital markets. The existing work focuses on collusive clique detections based on certain assumptions of trading behaviors. Effective approaches for analyzing and detecting wash trade in a real-life market have yet to be developed. This paper analyzes and conceptualizes the basic structures of the trading collusion in a wash trade by using a directed graph of traders. A novel method is then proposed to detect the potential wash trade activities involved in a financial instrument by first recognizing the suspiciously matched orders and then further identifying the collusions among the traders who submit such orders. Both steps are formulated as a simplified form of the knapsack problem, which can be solved by dynamic programming approaches. The proposed approach is evaluated on seven stock data sets from the NASDAQ and the London Stock Exchange. The experimental results show that the proposed approach can effectively detect all primary wash trade scenarios across the selected data sets.
A Novel Approach for Circular Trade Detection in Mercantile Exchange
The derivatives market having a significant number of investors trading in futures contracts, is vulnerable to manipulation by some perpetrators. Protecting market participants from a prevalent manipulation called circular trading and providing a fair market has always been a challenging task for regulators. This kind of malpractice is represented by the trading behaviors of a group of investors who trade among themselves frequently to increase the price of the commodity and consequently make forged prosperity. This paper presents a network-based approach for detecting investors involved in circular trading in the futures market. This is done initially by constructing the daily networks of investors' trades, then, extracting all trade cycles of various lengths from these daily networks to arrive at the group of initial suspicious cycle traders. Finally, in order to exclude investors who are randomly involved in suspicious cycles, price fluctuations over time were analyzed. The proposed approach has been conducted on real data from Iran Mercantile Exchange (IME) and as a warning system, has succeeded in detecting anomalous traders effectively.
Expert Systems with Applications, 2011
This paper addresses challenges relating to applying data mining techniques to detect stock price manipulations and extends previous results by incorporating the analysis of intraday trade prices in addition to closing prices for the investigation of trade-based manipulations. In particular, this work extends previous results on the topic by analysing empirical evidence in normal and manipulated hourly data and the particular characteristics of intraday trades within suspicious hours. Furthermore, the analytical models described in this paper reinforce the results of previous market manipulation studies that are based on traditional statistical and econometrical methods providing an alternative portfolio of methods and techniques originating from the data mining and knowledge discovery areas. With the application of the analytical approach described in this paper, it is possible to identify new fraud manipulation pattern characteristics encoded as decision trees which can be readily employed in fraud detection systems. The paper also proposes a number of policy recommendations towards increasing the effectiveness of the operational processes executed by stock exchange fraud departments and regulatory authorities.
The detection of market abuse on financial markets: a quantitative approach
Quaderni Di Finanza, No. 54, 2003
In every country with legislation on market abuse, i.e. on market manipulation and insider trading, the repression of these o¤ences is entrusted to supervisory and judicial authorities with powers that vary with the legislation in question. A procedure permitting cases of market abuse to be detected in real time is a need that is strongly felt by …nancial market supervisory authorities. Such a procedure consists basically in the analysis of the transactions carried out on the market by traders in order to detect anomalies that could be symptomatic of market abuse. The aim of this paper is to develop, through recourse to probability theory, a method for identifying cases of market abuse more e¤ectively.
Network-based Anomaly Detection for Insider Trading
2017
Insider trading is one of the numerous white collar crimes that can contribute to the instability of the economy. Traditionally, the detection of illegal insider trades has been a human-driven process. In this paper, we collect the insider tradings made available by the US Securities and Exchange Commissions (SEC) through the EDGAR system, with the aim of initiating an automated large-scale and data-driven approach to the problem of identifying illegal insider tradings. The goal of the study is the identification of interesting patterns, which can be indicators of potential anomalies. We use the collected data to construct networks that capture the relationship between trading behaviors of insiders. We explore different ways of building networks from insider trading data, and argue for a need of a structure that is capable of capturing higher order relationships among traders. Our results suggest the discovery of interesting patterns.
ANALYSIS OF STOCK MARKET MANIPULATIONS USING KNOWLEDGE DISCOVERY TECHNIQUES
This paper addresses challenges relating to applying data mining techniques to detect stock price manipulations and extends previous results by incorporating the analysis of intraday trade prices in addition to closing prices for the investigation of trade-based manipulations. In particular, this work extends previous results on the topic by analysing empirical evidence in normal and manipulated hourly data and the particular characteristics of intraday trades within suspicious hours. Furthermore, the analytical models described in this paper reinforce the results of previous market manipulation studies that are based on traditional statistical and econometrical methods providing an alternative portfolio of methods and techniques originating from the data mining and knowledge discovery areas. With the application of the analytical approach described in this paper, it is possible to identify new fraud manipulation pattern characteristics encoded as decision trees which can be readily employed in fraud detection systems. The paper also proposes a number of policy recommendations towards increasing the effectiveness of the operational processes executed by stock exchange fraud departments and regulatory authorities.
Data Mining in Financial Markets
The ongoing global financial recession has dramatically affected public confidence and market development. An example is the market manipulation schemes hidden in capital markets, which have caused losses in billions of dollars, dramatically damaging public confidence and contributing to the global financial and credit crisis. While most investors lost during market falls, for instance, sophisticated speculators can manipulate markets to make money by illegally using a variety of maneuvering techniques such as wash sales. With financial globalization, manipulators are becoming increasingly imaginative and professional, employing creative tactics such as using many nominee accounts at different broker-dealers. However, regulators currently are short on effective technology to promptly identify abnormal trading behavior related to complex manipulation schemes. As a result, shareholders are complaining that too few market manipulators were being caught. In this talk, I will discuss iss...
To Detect Irregular Trade Behaviors In Stock Market By Using Graph Based Ranking Methods
To detect the irregular trade behaviors in the stock market is the important problem in machine learning field. These irregular trade behaviors are obviously illegal. To detect these irregular trade behaviors in the stock market, data scientists normally employ the supervised learning techniques. In this paper, we employ the three graph Laplacian based semi-supervised ranking methods to solve the irregular trade behavior detection problem. Experimental results show that that the un-normalized and symmetric normalized graph Laplacian based semi-supervised ranking methods outperform the random walk Laplacian based semi-supervised ranking method.
A Graph Theoretical Approach for Identifying Fraudulent Transactions in Circular Trading
2017
Circular trading is an infamous technique used by tax evaders to confuse tax enforcement officers from detecting suspicious transactions. Dealers using this technique superimpose suspicious transactions by several illegitimate sales transactions in a circular manner. In this paper, we address this problem by developing an algorithm that detects circular trading and removes the illegitimate cycles to uncover the suspicious transactions. We formulate the problem as finding and then deleting specific type of cycles in a directed edge-labeled multigraph. We run this algorithm on the commercial tax data set provided by the government of Telangana, India, and discovered several suspicious transactions.
Bulk Classification of Trading Activity
SSRN Electronic Journal, 2000
The classification of the aggressor's side of a trade is a critical concern in Market Microstructure Theory. Among other uses, it is a key input necessary to identify information asymmetries and the presence of toxic order flow. Although some Exchanges have recently started to report the "aggressor" flag, historical databases usually lack this piece of data. Thus the researcher and/or practitioner still needs to infer the aggressor side from existing information, typically level 1 Tick Data. This poses the additional problem of having to parse hundreds of millions of records per instrument and year. In this paper we propose a new Bulk Volume Classification methodology that does not require working with Tick Data. Instead, it uses Time or Volume Bars, which for a small fraction of the records needed by the Tick rule delivers a classification with greater accuracy. The implication is that working with Tick Data for inferring the aggressor classification is not only inefficient and costly, but also does not offer greater accuracy compared to Time or Volume Bars.