基于非线性动力学的金融危机传染过程内生结构突变分析

发布时间：2018-05-10 06:44

本文选题：非线性动力学 + 金融危机传染　；参考：《哈尔滨工业大学》2014年硕士论文

【摘要】：20世纪90年代以来，频繁爆发的金融危机引起研究人员的广泛关注。这些危机都是在一个国家首先爆发，进而通过贸易、金融和预期等渠道传染到其他国家，对被传染国家和地区的金融市场和实体经济都造成了不可逆转的危害。历次金融危机的传染表现出强烈的系统性、复杂性和非线性特征，而传统的金融危机传染理论无法正确剖析金融危机传染期间的非线性特征。因此，本文尝试通过非线性动力学系统理论对金融危机传染进行研究，以此来刻画金融危机传染的非线性机制，以期在经济一体化、金融自由化的背景下，为我国的金融危机预警体系的建立提供合理化建议，维持金融体系的稳定。首先，本文从金融危机传染的理论、实证研究和非线性动力学在金融方面的应用两个方面综述该领域的国内外研究现状，然后介绍了基于非线性动力学时间序列预测的基础理论。在此基础上，，文本建立了基于非线性动力学的金融危机传染模型。将最大Lyapunov指数和非线性相互预测测度引入Bai和Perron的内生结构突变模型，构建了基于最大Lyapunov指数的传染源内生结构突变模型以及基于非线性相互依赖性的金融危机传染内生结构突变模型。然后，本文采用改进的基于最大Lyapunov指数的传染源内生结构突变模型对2008年金融危机期间传染源国家——美国的非线性动力学状态的变化进行实证分析。在仿真实验的基础上运用SP500股票指数数据进行实证分析，找到了金融危机每个时期的时间节点，对金融危机各阶段进行了划分与解释；比较了本文与采用Bai和Perron的内生结构模型的实证结果；分析了金融危机各阶段的非线性动力学特征的变化。最后，本文采用改进的基于非线性相互依赖性的金融危机传染内生结构突变模型分析2008年全球金融危机期间美国等10国（或地区）金融市场之间非线性相互依赖性的变动情况。首先采用Bai和Perron的内生结构突变模型对2008年金融危机数据进行验证分析。然后分别采用原始的非线性相互预测测度和基于非线性相互依赖性的金融危机传染内生结构突变模型对美国向其他9个被传染国家传染的方向、时间和强度进行分析。并在此基础上尝试为我国对金融市场危机的应对提供适用的工具及预警指标。
[Abstract]:Since the 1990 s, frequent financial crises have aroused the attention of researchers. These crises are the first to break out in one country, and then spread to other countries through trade, finance and expectation channels, causing irreversible harm to the financial markets and real economy of the infected countries and regions. The contagion of previous financial crises shows strong systematic, complex and nonlinear characteristics, but the traditional theory of financial crisis contagion can not correctly analyze the nonlinear characteristics of financial crisis contagion. Therefore, this paper attempts to study the contagion of financial crisis through the theory of nonlinear dynamics system, in order to describe the nonlinear mechanism of contagion of financial crisis, in the context of economic integration and financial liberalization. To provide reasonable suggestions for the establishment of financial crisis warning system and to maintain the stability of financial system. First of all, this paper summarizes the domestic and foreign research status of financial crisis contagion theory, empirical research and the application of nonlinear dynamics in finance. Then the basic theory of time series prediction based on nonlinear dynamics is introduced. On this basis, the text establishes the financial crisis contagion model based on nonlinear dynamics. The maximum Lyapunov exponent and nonlinear mutual predictive measure are introduced into the endogenous structural catastrophe model of Bai and Perron. An endogenous structural mutation model based on the maximum Lyapunov index and a nonlinear interdependency model for the endogenous structural mutation of the source of infection were constructed. Then, an improved model based on the largest Lyapunov exponent is used to analyze the change of nonlinear dynamics in the United States, the source country during the financial crisis in 2008. On the basis of the simulation experiment, the paper uses the SP500 stock index data to carry on the empirical analysis, has found the financial crisis each time node, has carried on the division and the explanation to the financial crisis each stage; The empirical results of this paper are compared with the endogenous structural models using Bai and Perron, and the changes of nonlinear dynamics characteristics in various stages of the financial crisis are analyzed. Finally, an improved financial crisis contagion endogenous structural catastrophe model based on nonlinear interdependence is used to analyze the nonlinear interdependence between the financial markets of the United States and other 10 countries (or regions) during the global financial crisis in 2008. Firstly, Bai and Perron's endogenous structural catastrophe model are used to verify and analyze the 2008 financial crisis data. Then we analyze the direction, time and intensity of the contagion from the United States to the other nine countries by using the original nonlinear mutual prediction measure and the endogenous structural catastrophe model of financial crisis contagion based on nonlinear interdependence. On this basis, it tries to provide applicable tools and early warning indicators for China's response to financial market crisis.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F224;F830.9

【参考文献】