当前位置: 首页>博士论文>资源详情
国内外金融市场间的相依结构及风险溢出关系研究
中文摘要

随着经济全球化与金融一体化的发展,国际金融市场间的联系日益增强,相依结构日趋复杂化、多元化。2007-2008年由美国次贷危机引发的全球金融危机,再次警醒着世界各国重新审视本国金融体系以及同国际金融市场间的关联关系。科学刻画金融市场间的相依结构及风险传导路径无论对于投资者的微观资产配置还是对于监管部门的宏观审慎管理和风险防范都有着重要的现实价值。 基于此,本文以国内外主要金融市场作为研究对象,揭示其相依结构和风险溢出关系。从技术角度而言,金融市场间相依性的准确描述在很大程度上依赖于金融建模工具,其中如何测度多市场之间的相依性是一个核心问题,这对于风险评估和监管而言尤为重要。然而,长期以来金融建模大都考虑多元正态分布,并假设资产收益率的联合分布和边缘分布均服从正态分布。大量的实证研究却表明金融资产收益率分布是非正态的,它具有显著的尖峰、厚尾、有偏等“典型事实”特征,多资产间的相依性也表现出显著的尾部相依性以及非线性和非对称性相依性等特征。经典的多元正态分布假定不能体现以上特征性事实,无法满足精确刻画金融市场间复杂相依结构的需要。Copula函数能够较好地解决这个问题,它可以度量非线性或非对称相依性,能够较好地刻画资产间的复杂相依结构。 鉴于此,本文基于Copula函数对国内外金融市场间的相依性做了系统的研究。不仅如此,为了揭示金融市场间风险溢出的程度、方向以及极端风险溢出,本文还借助于Diebold & Yilmaz溢出指数模型和尾部相依性测度对国内外金融市场间的风险溢出关系进行了深入的分析。 具体而言,本文系统梳理了金融市场间的相依性和风险溢出测度,总结了各测度的优点和不足,揭示了金融市场相依性和风险溢出的内在机制。实证研究层面,本文从收益率相依结构、波动率溢出关系、极端风险溢出等角度全方位揭示了国内外金融市场间的关联关系的全景图。具体的研究工作如下: (一)研究了国内外金融市场间收益率的相依结构。本文将DCC-GARCH模型和Copula函数相结合,得到一个新的模型Copula-DCC-GARCH模型,并运用该模型研究了国内外股票市场间的相依性。为更细微的观察国内外股票市场间相依性的时变性特征,运用结构变点检测方法对国内外股市的动态条件相关系数进行了结构变点检测。实证研究结果表明(1)总体上而言,国内外股票市场间的相依性是逐步提升的:(2)国内外股市间的相依性具有时变性,并存在3-4个变点,这些特征与国内外股市间的经济金融基本面因素相关。(3)从国别结构上看,我国股市与亚洲股市的相依性最为紧密,其次是金砖国家,与欧美股市的相依性较弱。 (二)揭示了国内外金融市场间的波动率溢出关系。为了刻画国内外金融市场间波动率溢出的强度和方向,本文运用Diebold&Yilmaz溢出指数模型研究了国内外股市间波动率的溢出关系。研究结果表明:(1)总体说来,国内外股票市场的波动率的溢出效应大于收益率的溢出效应,且收益率和波动率的对外总溢出效应都存在时变性和非对称性。(2)欧美发达国家股市对其它股市的溢出效应要大于其它股市对它们的溢出效应,是溢出的净传输者,相反的,韩国和日本以及金砖国家股市是溢出的净接受者,这体现了发达国家等成熟股市开放水平高,在国际股市中占居重要影响力。(3)波动率溢出效应具有聚集特征,在金融危机期和经济平稳期表现各不同,在危机期波动溢出效应大,2008年全球金融危机时期,国际股市波动溢出效应出现了跳跃性的上升。 (三)研究了国内外金融市场间的极端风险溢出关系。本文首先用三种不同的Vine Copula模型刻画了国内外股市间的整体相依结构特征,然后用尾部相依性测度考察了两两股市间的风险溢出关系,研究结果表明:(1)国际股市间的相依结构呈现一定的经济体集聚特征,采用不同的Vine Copula模型其相依结构略有差异;(2)国际股市间存在非对称的尾部相依性,且下尾相依性普遍高于上尾相依性;(3)成熟市场对外的尾部相依性均值较高,新兴市场的较低,说明发达国家的金融市场在国际金融市场中仍占主导地位。 本文侧重于实证研究,与以往的研究相比,本文的研究更系统、全面。本文分别从收益率相依结构、波动率溢出关系和极端风险溢出关系多个角度对全球主要金融市场间的关联关系进行了研究。本文还基于Vine Copula方法,刻画了国内外金融市场间的整体的相依结构及风险的传导路径。 关键词:Copula函数;Vine Copula函数;相依性;时变性;风险溢出

英文摘要

With the development of economic globalization and financial integration, the links between international financial markets have become increasingly stronger, and the dependent structure has become increasingly complex and diversified. The global financial crisis triggered by the subprime mortgage crisis in the United States in 2007-2008 once again awakened the countries of the world to re-examine their own financial systems and their relationship with international financial markets. Scientifically characterizing the dependent structure and the path of risk transmission in international financial markets has important practical value both for investors' micro-asset allocation and for macro-prudential supervision and risk prevention of regulatory authorities. Based on this, the paper uses domestic and foreign major financial markets as research samples to reveal its correlation structure and risk spillover relationship. From a technical point of view, an accurate description of the interaction between financial markets relies heavily on accurate financial modeling tools, and one of the key issues is how to measure the interdependencies among multiple assets, which is particularly important for risk assessment and regulation. However, for a long time, financial modeling has mostly considered multivariate normal distribution, assuming that the joint distribution and marginal distribution of return on financial assets follow a normal distribution. A large number of empirical studies show that the distribution of return on financial assets is non-normal, and it has obvious "typical fact" features such as high peak, fat tail and bias. The interdependency among multiple assets also shows significant tail dependence as well as non-linear and asymmetric characteristics. The classical multivariate normal distribution assumption can not reflect the above characteristics and can not meet the need of accurately describing the complicated and dependent structure between financial markets. The Copula function can solve this problem well; it can measure the nonlinear or asymmetric dependencies, and can flexibly construct the dependency structure between assets. In view of this, the paper makes a systematic study of the interdependence between domestic and international financial markets based on the Copula function. Moreover, in order to reveal the extent and direction of risk spillovers among financial markets and the spillover of extreme risks, this paper also uses the Diebold & Yilmaz spillover index model and tail dependency measure to analyze the risk spillover relationships between domestic and foreign financial markets. Specifically, this paper systematically sorts out the interdependence and risk spillover measures between financial markets, summarizes the strengths and weaknesses of each measure, and reveals the internal mechanisms of financial market interdependence and risk spillover. At the level of empirical research, this paper comprehensively reveals the overall picture of the correlation between domestic and foreign financial markets from the aspects of return dependent structure, volatility spillover relationship, and extreme risk spillover. Specific studies works are as follows: First, it studies the return dependent structure between domestic and foreign financial markets. This paper combines the DCC-GARCH model and Copula function to get a new Copula-DCC-GARCH model, and use this model to study the dependence of domestic and foreign stock markets. In order to further observe the time-varying characteristics of the dependence of domestic and foreign stock markets, this paper uses the structural change detection method to detect the structural change points of the dynamic condition correlation coefficient between domestic and foreign stock markets in the world. The empirical results show that: (1) Generally speaking, the dependence of domestic and foreign stock markets is gradually increasing; (2) The dependence of domestic and foreign stock markets is time-varying, and there are 3-4 change points, which are related to the economic and financial fundamentals between domestic and foreign stock markets. (3) From the perspective of country structure, China's stock market and the Asian stock market are the most closely linked, followed by the BRICS countries, and the dependence on European and American stock markets is weak. Secondly, it reveals the volatility spillover relationship between domestic and foreign financial markets. In order to characterize the intensity and direction of volatility spillovers between domestic and foreign financial markets, this paper uses the Diebold & Yilmaz spillover index model to study the spillover relationship between domestic and foreign stock market volatility. The results show that: (1) Generally speaking, the spillover effect of volatility in the domestic and international stock markets is greater than the spillover effect of returns, and the total external spillover effect of returns and volatilities has time-variability and asymmetry. (2) Developed countries in Europe and the United States have more spillovers over other markets than other stock markets to them, which are net transmitters of spillovers. Conversely, South Korea, Japan and BRICS stock markets are net receivers of spillovers. This shows that mature markets such as developed countries have a high level of openness and occupy an important influence in the international stock market. (3) The volatility spillover effect has the characteristics of aggregation, which shows different performance during the economic crisis and economic stabilization, with large volatility spillover effect during the crisis period. During the global financial crisis in 2008, the volatility spillover effect of the international stock market showed a jumpy increase. Thirdly, it studies the extreme risk spillovers between domestic and foreign financial markets. This paper firstly uses three different Vine Copula models to characterize the overall dependence structure of the international stock market. Then, the tail dependence measure is used to examine the risk spillover between each two stock markets. The results show that: (1) The dependent structure of international stock markets presents some characteristics of economic agglomeration, and the dependent structure is slightly different with different Vine Copula models. (2) There is asymmetrical tail dependence between international stock markets, and the bottom tail dependence is generally higher than the upper tail dependence. (3) The average value of external tail dependence in mature markets is relatively high, while that is low in emerging markets, indicating that the financial markets of developed countries still dominate the international financial markets. This paper focuses on empirical research. Compared with previous studies, the research in this paper is more systematic and comprehensive. This paper studies the relationship between the major global financial markets from the perspectives of yield-dependent structure, volatility spillover relationship, and extreme risk spillover relationship. Based on the Vine Copula method, this paper also describes the overall interdependent structure and risk conduction path between domestic and foreign financial markets. Keywords: Vine Copula; dependence; time-variability; risk spillover

作者相关
主题相关
看过该书的人还在看哪些书