Research On Agent-based Stock Exchange

This paper gives picture of research to stock exchange system analysis using agent-based model with Swarm simulation platform. An agent-based artificial stock market modeling is from the bottom up with multiple interacting agents. Santa Fe Institute artificial stock market is a well known agent-based model and one of the benchmark for researchers to study. This artificial stock can be securitized from the complex system perspective and examines how the interaction between traders the microscopic level can result in macroscopic behavior. In spite of the model is built on basis of the mathematical function. The innovation lies in the paper of the newly-borne multi-agent simulation methodology with computer science.The paper establish an artificial simulation model, which is based on the financial market microstructure theory, behavioral finance theory, and multi-agent simulation technology on Java-Swarm ASM 2.2.3. Swarm is a programming library, it represents an attempt to gather up many different kinds of models that go under the heading of "agent-based modeling" and create a common language and programming approach. We define the mathematical model such as agent's expectation, requirement, rule, evolution and world. The software framework use UML design method. The system utilize genetic algorithm simulate a leaning pool to make agents having the capability of self-leaning. Agents who are publishers may publish rules into learning pool to share with others; receivers may retrieve better rules from learning pool. The simulation parameters are designed and the interaction between agents are simulated to observe the dynamic,evolutionary process of organization formation by GUI and analyze the results of the simulation.I observe the individual learning mechanism, and prove that the market is a rational one compared with realistic market. I can observe more features common to real markets in this virtual market. The simulation results indicate that the model is effective.