Research On The Multivariate Conditional Dependent Structural Breaks Identification And Application Of Dependent Measure

As the tool to study the construction of multivariate joint distribution function,the construction of multivariate dependence structure and the measurement of dependence,Copula Function Theory can not only fully explore the data information of a single variable,flexibly construct multivariate joint distribution function,but also realize the construction of multivariate dependent structure at the same time,which can effectively capture the characteristics of nonlinearity,asymmetry and weak tail correlation among variables,and has good data fitting ability.With the introduction of Vine,a graphic tool,the Copula function theory realizes the conditional dimension reduction of multivariate cases,that is,Vine structure is used to achieve the conditional degradation of multivariate complex problems,split into a number of binary cases,and on this basis,the corresponding conditional processing is carried out.Through Vine’s "conditional dimension reduction",the Copula function theory is effectively applied in high-dimensional scenarios.Vine Copula Model can also realize the construction of conditional dependence structure and the measurement of conditional dependence,which greatly expands the research scope of Copula Function Theory,and the importance and application value of Copula Function Theory and Vine Copula Model are revealed.Under the current economic and financial globalization background,Vine Copula Model can be fully used to construct and measure the conditional dependence structure and dependence among international financial markets,and to find the associated risks among financial markets in a timely manner.Therefore,it has certain theoretical significance and application value to continue to expand the research on Vine Copula Model.As the core of modern economy,finance plays an important role in the construction of social credit system,the allocation of economic resources and high-quality development.At present,with the continuous strengthening of economic and trade exchanges,capital flows and information transmission,international financial markets are gradually deeply integrated,developed and evolved.The interdependent structure of financial markets is characterized by variable structure.Financial risks are constantly cross-infected among markets,and the spillover capacity of volatility risks is gradually increasing.And by the overtures of Sino-US Trade Friction,"The Global COVID-19" and "the conflict between Russia and Ukraine" superimposed effect of the factors such as the era background,accurate to build the international financial market condition dependent structure and measure the dependency between,do financial risk contagion path analysis is not only the premise and key volatility spillover and risk measurement.At the same time,it is also the research focus of various financial market regulators,financial practitioners and scholars.Therefore,it is very necessary to use the Vine Copula model with "robustness" to study the conditional dependent structure of financial markets.This thesis systematically combs the latest research results of Copula Function Theory,and finds that the current Vine Copula Model has some practical application effects,but the research on its robustness is not perfect.Although some scholars have begun to study the change point of Vine Copula Model parameters,they have not further explored the changes of Vine and PairCopula function types in the model.In fact,Vine and Pair-Copula function types and their parameters in multivariate conditional dependency structure are mutually affected.Therefore,this thesis uses the modified ICSS algorithm and the idea of likelihood ratio test,do the research of the robustness content of Vine Copula model.The innovation point of this thesis mainly has three points:(1)The inductive classification of Copula function is supplemented and improved,the classification of Copula model is extended,and the classification standard of Copula function is optimized.On the basis of the existing classification criteria of Copula function,combined with the application characteristics of multivariate scenarios,this thesis further clarifies the classification criteria of Copula model,which is convenient for the continuous expansion and application research of Copula function theory in multivariate scenarios.(2)On the basis of the construction of Vine Copula model,the research content of structural break identification is added to improve the modeling efficacy of the model and the fitting ability of the data.Based on the Vine Copula model,the thesis carries out relevant research on the identification of structural breaks,and integrates the identification of volatility structure breaks and the identification of conditional dependence structure breaks driven by data.On the basis of theoretical analysis and computer random simulation,the validity of the proposed structural break identification research method is cross-verified.This enriches the research on the robustness of Vine Copula model and proposes a feasible technical research path for the mutation identification of multivariate conditional dependence structure.(3)By using the mutation identification method of multivariate conditional dependence structure,this thesis conducts an empirical study on the conditional dependence structure of major international stock markets,and finds the potential variable structure characteristics and segmented dependence characteristics.The empirical analysis shows that the conditional dependence structure of the major international stock markets has significant variable structure characteristics,and the Kendall rank correlation coefficient and the upper and lower tail correlation coefficients are adjusted in sections.In addition,the stock markets of France and Germany are the "core nodes" of the conditional dependence structure among international financial markets,and other financial markets are directly or conditionally related to them,with certain risk dependence among them.Although China’s stock market is at the "edge" of the conditional dependence structure,we should continue to pay attention to the external risk contagion and the internal upper and lower tail dependence of the stock markets of France,Germany,Japan and the Unted States,so as to find the sudden change of the conditional dependence structure in time and do a good job in the basic research for the prevention and control of China’s external financial risk contagion and the measurement of volatility spillover.The thesis adopts the method of theoretical analysis and computer stochastic simulation to carry out the research of multivariate conditional dependent structural mutation identification.In the research of the identification part of fluctuation structure mutation,the effect of fluctuation structure mutation on multivariate conditional dependent structure is analyzed by combining stochastic simulation and real data.However,in the research of multivariate conditional dependent structural break recognition,it is completely based on theoretical analysis and computer stochastic simulation.As a whole,this thesis provides a feasible technical research path for multivariate conditional dependence structural mutation identification and enriches the research content of the robustness of Vine Copula Model.