Partition Construction Of Copula Function And Its Application

The dependence relationship between random variables has important research value in practical applications,Although covariance matrices are still widely used,they can only characterize the linear dependencies between variables.Because the Copula correlation theory can better describe the relationship between the joint distribution of selected variables and the marginal distribution corresponding to the variables,it is widely used to characterize the dependence between variables.At present,the commonly used Copula function models mainly include : R-Vine-Copula,Normal Copula function,t-Copula function,and Archimedes function Copula function,but in the study of complex nonlinear dependencies,these models cannot comprehensively used to describe the linkage between variables.Based on the above situation,this paper presents a method for constructing a Copula model,which is abbreviated as: S-C(Slice-Copula)function.The specific construction steps are as follows:(1)Slice the domain of the Copula function [0,1]×[0,1],this article cuts the binary sample set into five slices.(2)Select the corresponding Copula function for each slice;(3)Estimating the parameters of the selected Copula function;(4)Coupling the selected Copula function with the method of OLS;As an example,we used the newly constructed S-C function to study the dependence characteristics between the Shanghai Stock Index and Fiberhome's daily returns.The empirical results show that: using the S-C function model based on the shard structure to construct the Shanghai Stock Index and Fiberhome When communicating the joint distribution function between the daily returns of the two stocks,the results obtained are more accurate;at the same time,the two selected stocks also have an asymmetric tail relationship.In addition,this paper also proposes a VaR model based on the S-C function,and through comparative analysis it is concluded that this model can better calculate the VaR under the equal weight combination of the Shanghai Stock Index and FiberHome.