Dissertation > Excellent graduate degree dissertation topics show

Study on the Improved Modelof Artificial Stock Market Based on Agent and Scale-Free Network

Author: SunHongZuo
Tutor: SunBaiQing
School: Harbin Institute of Technology
Course: Finance
Keywords: Artificial stock market Agent Scale-free networks Complexity
CLC: F224
Type: Master's thesis
Year: 2011
Downloads: 70
Quote: 0
Read: Download Dissertation

Abstract

With the depth and breadth of complexity science is rising, more and more scholars and researchers introduced the concept of complexity to the study of the financial market. As a complex giant system, the complexity of the characteristics of the financial markets has its own unique performance characteristics such as excessive volatility of the yield, fat tail, fractal and volatility clustering effect. These complex features found shaken the theoretical basis of the efficient market hypothesis, the traditional finance research paradigm presented a huge challenge. In order to explore the underlying causes of these complex characteristics, in order to develop more in line with reality theories and models of financial markets, a combination of computer technology, cross-discipline of complexity science and behavioral finance theory sprung up, which is based on the calculation of Agent Economics (ACE). By the bottom-up modeling approach, the use of high-speed information processing capabilities of the computer, establish an artificial stock market, highly realistic simulation of the real financial markets, can be used to guide the financial market regulators and policy makers \; addition, the study of the impact of infection mechanisms in individual decision-making mechanism of the artificial stock market investors and groups, not only conducive to reveal the asset price discovery and formation mechanism, but also for the improvement of the market structure and market-oriented operation mechanism, strengthen market supervision institution-building, regulate the behavior of traders, to protect the interests of investors, and has a positive meaning. First at home and abroad Research on Agent-based artificial stock market modeling are reviewed, on the basis of the theory, including complex scientific theory, the theory of scale-free networks and Agent-based modeling method. Build artificial stock market improved model of bounded rationality of investors and heterogeneity of this basic problem, and data on the investments of investors of different countries and regions rational differences, and then put forward the improvement of the basic assumptions of the model . Model by introducing the scale-free network topology of the network has been improved, and accordingly design the Agent interaction coefficient, so that the investors' decision-making function more in line with reality. The combination of investors to determine the fundamental value of the stock and the market pricing mechanism, this paper Matlab programming model simulation. Simulation results show that: the scale-free network topology is more in line with the actual stock market relationships; investors as the \mechanism designed to be explained. Established in this paper to improve artificial stock market model to some extent to achieve a simulation of the real stock market, which indicates that Agent and scale-free networks based on improved artificial stock market modeling method is effective, and thus the value of the model analysis also has important practical significance and quite targeted.

Related Dissertations

CLC: > Economic > Economic planning and management > Economic calculation, economic and mathematical methods > Economic and mathematical methods
© 2012 www.DissertationTopic.Net Mobile