Non-linear Evolution Research For Stock Market,Money Market And Foreign Exchange Market

Stock market, Money market and Foreign exchange market are the most important factors of the financial market. The healthy development of them are closely related to the economic development of our country and the stability of financial market. Three markets are not ndependently operated, but rather connected to each other, and promote each other. With the development of the domestic financial market, the relations among them bring out characteristics of nonlinearty, complexity and time-varying. However, while researchers have not yet provided a satisfying answer to the problem of relationships among the three markets, the problem is still in needs of an effective theory and study system. For this situation, the article which explores the appropriate models and methods of the nonlinear evolution research for stock market, money market and foreign exchange market based on the consideration of macroeconomic fundamentals was studied, this paper studies from the following aspects:First of all, this paper established alternative library of indexes for the three financial submarkets and macroeconomic indicators based on the economic implication and existing research. Alternative library contains several important indexes on the market, covering most of the information of the markets. This paper used Granger causality test and nonlinear causality test to judge the causal relationships between indicators. Then we selected some market indexes which interact with others obviously.Next, market indexex are used to establish comprehensive indexes of the stock market, money market, foreign exchange market and macroeconomic, to ensure comprehensive indexes contain most of information of the markets and the causal relationship of different markets comprehensive indexes are obvious. Considering the influence of the macroeconomic, this article proposed the nonlinear evolution model among the three submarkets based on ordinarydifferential equations. Considering characteristic of complexity and time-varying on financial market produce a lot of noise information, this paper join cardinality constrain in the original model to reduce noise components of the original model. Due to the conventional parameter estimation methods can’t effectively solve problems with lots of parameters and cardinality constraint, this paper introduced an intelligent optimization algorithm called COMDE as a parameter estimation method.Then, this paper made empirical researches on the relationship of nonlinear evolution between two or among three financial submarkets with the comprehensive indexes and nonlinear evolution models. Empirical results show that the nonlinear evolution model with constrains can well depict the nonlinear evolution of the three markets’ structure and reveal the interaction relationship among these markets. This model has the advantages of high precision and intelligibility.At last, to verify the applicability of the intelligent omptimization algorithm in financial field, this paper used DE-ELM forecasting model to forecast tendency of Shanghai Composite Index and S&P 500 Index. Empirical results show that intelligent optimization algorithms can effectively used in financial field, them provides the feasible investment decision supports for the investors.