Study On Financial Complexity And Modeling Financial Market

Finance, a crucial component of the world's economy, plays a more and more important role in the development of economy. However, the traditional financial theories, which have a reign of the past 40 years in describing the behaviors of price fluctuations, are discovered great distinction with that of the real market. A large amount of "abnormal phenomena" have been found since 1980s, inconsistent with the traditional theories, which have shaken the fundamental of financial theories formed under the linear uniform frame — the Efficient Market Hypothesis. The enigma of the "abnormal phenomena", the underlying mechanism as well as the shortcomings of the financial theory system, attract great many researchers dedicating themselves to investigate rules of capital markets with new nonlinear methods, as an opportunity for the complex science flourishing.I also dedicate myself into such field as mentioned above. When dealing with the complex financial problems, the market is considered as a nonlinear dynamic system, a model based on self-organized percolation is constructed and the behaviors of price fluctuations and market evolvement are investigated, adopting new methods rather than the traditional ones. As a result, several productions of good theoretical values and practical meanings are as follows.· A statistical method to investigate the return distribution with variable location and time scale measurement - the Continuous Avalanche Model is proposed, while the distribution of the continuous avalanche size of the time series of the real market index is analyzed. Furthermore, the power-law distribution on the tail of them is discovered. Return distribution analyzed by the above method is not only in accord with the practical system, but also uncovers the correlation between price fluctuations, which is conducive to understanding the market dynamical evolvement.· A model based on self-organized percolation is established. It ingeniously coevolves with the 'herding' behavior, the nonlinear interactions between agents and the variance of the systemic nonlinear structure, successfully...