A Hidden Markov Model with Abnormal States for Detecting Stock Price Manipulation

Yi Cao; Yuhua Li; Sonya Coleman; Ammar Belatreche; Martin McGinnity

doi:10.1109/SMC.2013.514

A Hidden Markov Model with Abnormal States for Detecting Stock Price Manipulation

Yi Cao, Yuhua Li, Sonya Coleman, Ammar Belatreche, Martin McGinnity

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

12 Citations (Scopus)

Abstract

Price manipulation refers to the act of using illegaltrading behaviour to manually change an equity price withthe aim of making profits. With increasing volumes of trading,price manipulation can be extremely damaging to the properfunctioning and integrity of capital markets. Effective approachesfor analysing and real-time detection of price manipulation areyet to be developed. This paper proposes a novel approach,called Hidden Markov Model with Abnormal States (HMMAS),which models and detects price manipulation activities. Togetherwith the wavelet decomposition for features extraction andGaussian Mixture Model for Probability Density Function (PDF)construction, the HMMAS model detects price manipulationand identifies the type of the detected manipulation. Evaluationexperiments of the model were conducted on six stock tick datafrom NASDAQ and London Stock Exchange (LSE). The resultsshowed that the proposed HMMAS model can effectively detectprice manipulation patterns

Original language	English
Title of host publication	Unknown Host Publication
Publisher	IEEE
Pages	3014-3019
Number of pages	6
DOIs	https://doi.org/10.1109/SMC.2013.514
Publication status	Published (in print/issue) - Oct 2013
Event	IEEE International Conference on Systems, Man, and Cybernetics - Manchester Duration: 1 Oct 2013 → …

Conference

Conference	IEEE International Conference on Systems, Man, and Cybernetics
Period	1/10/13 → …

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/SMC.2013.514

Cite this

@inproceedings{8923e277f8e64cb5988925642755798e,

title = "A Hidden Markov Model with Abnormal States for Detecting Stock Price Manipulation",

abstract = "Price manipulation refers to the act of using illegaltrading behaviour to manually change an equity price withthe aim of making profits. With increasing volumes of trading,price manipulation can be extremely damaging to the properfunctioning and integrity of capital markets. Effective approachesfor analysing and real-time detection of price manipulation areyet to be developed. This paper proposes a novel approach,called Hidden Markov Model with Abnormal States (HMMAS),which models and detects price manipulation activities. Togetherwith the wavelet decomposition for features extraction andGaussian Mixture Model for Probability Density Function (PDF)construction, the HMMAS model detects price manipulationand identifies the type of the detected manipulation. Evaluationexperiments of the model were conducted on six stock tick datafrom NASDAQ and London Stock Exchange (LSE). The resultsshowed that the proposed HMMAS model can effectively detectprice manipulation patterns",

author = "Yi Cao and Yuhua Li and Sonya Coleman and Ammar Belatreche and Martin McGinnity",

year = "2013",

month = oct,

doi = "10.1109/SMC.2013.514",

language = "English",

pages = "3014--3019",

booktitle = "Unknown Host Publication",

publisher = "IEEE",

address = "United States",

note = "IEEE International Conference on Systems, Man, and Cybernetics ; Conference date: 01-10-2013",

}

TY - GEN

T1 - A Hidden Markov Model with Abnormal States for Detecting Stock Price Manipulation

AU - Cao, Yi

AU - Li, Yuhua

AU - Coleman, Sonya

AU - Belatreche, Ammar

AU - McGinnity, Martin

PY - 2013/10

Y1 - 2013/10

N2 - Price manipulation refers to the act of using illegaltrading behaviour to manually change an equity price withthe aim of making profits. With increasing volumes of trading,price manipulation can be extremely damaging to the properfunctioning and integrity of capital markets. Effective approachesfor analysing and real-time detection of price manipulation areyet to be developed. This paper proposes a novel approach,called Hidden Markov Model with Abnormal States (HMMAS),which models and detects price manipulation activities. Togetherwith the wavelet decomposition for features extraction andGaussian Mixture Model for Probability Density Function (PDF)construction, the HMMAS model detects price manipulationand identifies the type of the detected manipulation. Evaluationexperiments of the model were conducted on six stock tick datafrom NASDAQ and London Stock Exchange (LSE). The resultsshowed that the proposed HMMAS model can effectively detectprice manipulation patterns

AB - Price manipulation refers to the act of using illegaltrading behaviour to manually change an equity price withthe aim of making profits. With increasing volumes of trading,price manipulation can be extremely damaging to the properfunctioning and integrity of capital markets. Effective approachesfor analysing and real-time detection of price manipulation areyet to be developed. This paper proposes a novel approach,called Hidden Markov Model with Abnormal States (HMMAS),which models and detects price manipulation activities. Togetherwith the wavelet decomposition for features extraction andGaussian Mixture Model for Probability Density Function (PDF)construction, the HMMAS model detects price manipulationand identifies the type of the detected manipulation. Evaluationexperiments of the model were conducted on six stock tick datafrom NASDAQ and London Stock Exchange (LSE). The resultsshowed that the proposed HMMAS model can effectively detectprice manipulation patterns

U2 - 10.1109/SMC.2013.514

DO - 10.1109/SMC.2013.514

M3 - Conference contribution

SP - 3014

EP - 3019

BT - Unknown Host Publication

PB - IEEE

T2 - IEEE International Conference on Systems, Man, and Cybernetics

Y2 - 1 October 2013

ER -

A Hidden Markov Model with Abnormal States for Detecting Stock Price Manipulation

Abstract

Conference

UN SDGs

Access to Document

Fingerprint

Cite this