Babis Theodoulidis
University of Manchester - Manchester Business School

David Diaz
University of Manchester - Manchester Business School; Universidad de Chile - Escuela de Economia y Negocios

Abstract

Market monitoring is a very important process to the stock market. It is necessary in order to verify that all trades comply with the existing rules, as well as to detect any act of manipulation. Recently, a new kind of trading has emerged, High Frequency Trading, which allows traders to place and execute orders within milliseconds via a program running in a computer. It is doubtful whether existent market systems are capable of detecting inconsistencies in trades at this speed. This project has attempted to propose a design of a detection engine that could be incorporated to a monitoring framework so as to accommodate High Frequency trades. Understanding the field of stock markets and High Frequency Trading was a crucial part of this project. Since it is such a recent phenomenon with no confirmed cases (to the best of our knowledge) of market manipulation, we attempted to answer whether there are certain conditions that could benefit market manipulators. We used business intelligence techniques to analyse historical data and discover what sort of indications our detection engine should look for. Our results show that there have been violations of regulations that were not blocked in real-time which proves the inefficiency of current market monitoring systems. We also prove that an extreme number of orders within seconds can delay an exchange’s processes and systems. In the end we propose a design that takes those results into consideration.