Development of an Agent-Based Market Simulator

Financial markets today are monitored and controlled by artificial intelligent algorithms. Developing these artificial intelligent algorithms requires a large amount of testing against complex patterns and phenomena observed in stock market. The main objective of the proposed project is the development of a market simulator. This is highly challenging and yet promising direction that will allow the partner to test some of its algorithms locally before running them in the live market. The market simulator is proposed to have an exchange engine and computerized agents that will be interacting just like participants in the real market. Rationale of the agents can be derived by observing real market entities and algorithmically extracting valid elements of their behavior. By adjusting certain parameters of the agents, one can create very specific conditions or scenarios to optimize the performance of the algorithms.

Faculty Supervisor:

Dr. Stephen Watt

Student:

Rui Hu

Partner:

Cyborg Trading Systems

Discipline:

Computer science

Sector:

Finance, insurance and business

University:

Western University