Financial markets abuse activities aim to artificially manipulate prices or trading volumes of financial instruments to make a profit. Among three primary market abuse formats, information-based, action-based and trade-based manipulation, trade-based manipulation has attracted much attention from regulators around the world especially since the flash crash in 2010. This talk presents novel methods for detecting trade-based market manipulations, specifically price manipulation and volume manipulation.

Two detection models, static model and dynamic, are proposed for detecting different aspects of the price manipulation activities. The static model considers each trading action as a single object without contextual relations and uses a non-stationary feature reduction transformation together with classifiers as the detection model. The dynamic model considers the temporal contextual relationship among sequential trading behaviours and uses a hidden Markov model-based algorithm to detect complex price manipulation behaviours. This talk also proposes a directed graph and dynamic programming-based two-step algorithm for detecting wash trade, a major format of volume-based manipulation. 

All the proposed models are extensively evaluated against existing models on various datasets, including synthetically generated datasets and real market datasets donated by collaborators of financial companies. Experimental results demonstrate that the Static Model and Dynamic Model can effectively detect price manipulation activities, i.e. spoofing trading and quote stuffing and outperform the selected benchmark models; the directed graph and dynamic programming-based two-step algorithm on wash trade detection is also effective under different parameter configurations.

Dr Yuhua Li is a senior lecturer at the School of Computer Science and Informatics, Cardiff University, where he leads the Data Analytics and Machine Learning Research Group. His research covers fundamental and applied research in artificial intelligence and machine learning. In particular, he has made contributions to several key AI and machine learning topics, including novelty/anomaly detection, data reduction (including pattern selection and feature selection), learning from imbalanced data, dataset drift detection, neural networks/deep learning, explainable artificial intelligence, hyperdimensional computing, semantic similarity analysis, and forecasting. His work has motivated other researchers to develop new algorithms, use them as benchmarks and adopt them in products. He has applied machine learning to solve FinTech problems in stock markets, lending and insurance.