Presentation - Stock Market Layering Fraud as a Stochastic Optimal Stopping Problem

The proliferation of algorithmic high-frequency trading in financial markets has also led to an increase in new types of fraudulent activity. Since the flash-crash of 2010 first brought it to popular prominence, layering or spoofing fraud has become a major concern for financial regulators around the globe. Research into layering fraud has followed two general approaches: (i) an empirical analysis of layering fraud cases and the accompanying deviation in market variables, and (ii) detection studies based on the pattern matching with known fraud cases. In the present study we take the novel perspective of approaching layering fraud as an optimization problem, based on stochastic control theory. ***Our motivation is to provide researchers and regulators with the foundation of a general analytic framework to understand and model layering fraud, and to potentially detect new (unseen) patterns of layering across stocks and stock markets.***