Simulation And Research On The Capital Flow Of Security Market In The Financial System

Financial system is a complex giant system which is difficult to simulate and study directly for its non-linearity, high coupling and self-organizing. However, we could firstly study the pattern of individual behavior in the financial system to reflect the macroeconomy by the group behavior of individuals. And then a corresponding parallel system can be built to make a series of experiments, in order to study the pattern of the financial system and solve the untestable problem in the real life social financial system.The paper makes a description of the security market at first, and analyse the attribute sets and event lists of all the agents in the security market, including retail investors, institutional investors, security companies, the government and the stock exchange. Then the stock market and bond transaction market are made, with establishinhg the investment strategy and rule base of all the investors and the trading mechanisms of both markets.Then, by agent-based modeling technology, the agents of family, enterprise, bank and government are made with all the core components like attribute sets, event lists and rule bases. The paper studies the information flow and capital flow among all the agents, and establishes a simulation model of the security market in the financial system.Finally, a simulation platform is developed and the system is based on Java. And a series of experiments are made on the platform, such as studying the simulation results under different macroeconomic policies. Then the paper compares the results with traditional economic theories.In conclusion, the financial simulation model is analysed and the platform is established which lay a critical stone on the future financial simulation research.