Michal Galas, Dan Brown and Philip Treleaven

This paper describes a major computational social science environment for conducting experiments in finance and economics, specifically for investigating trading algorithms and economic risk.

For the past seven years UCL has worked with the major investment banks developing algorithms for trading systems, and with the regulators investigating high- frequency trading risk and systemic risk. To support this work we have developed two computational environments/platforms: a) our Algorithmic Trading & Risk Analytics Development Environment (ATRADE), and b) our social media streaming, storage and analytics platform (SocialSTORM).

This paper focuses on algorithmic trading (i.e. ATRADE) designed specifically to study the behavior and risk of trading algorithms. In the paper we briefly describe the algorithms used for automated trading and a library of algorithms we have developed which we believe is unique in academia. In addition, ATRADE is used to support an annual global algorithmic trading competition which provides a basis for evaluating performance.

Download PDF.