Copula函数选择的小波方法  

Copula Selection through Wavelet Methods

作　　者：彭选华[1] 

机构地区：[1]西南政法大学经济学院,重庆401120

出　　处：《西南大学学报（自然科学版）》2016年第8期90-99,共10页Journal of Southwest University(Natural Science Edition)

基　　金：重庆市教委科学技术研究项目(KJ130107);重庆市自然科学基金项目(cstc2012jjA00023)

摘　　要：金融资产相依结构研究中选择Copula函数很关键.考虑到相依结构的局部特征差异,利用小波函数的局部自适应能力,将阈值规则引入Copula理论,提出Copula函数的小波收缩估计量,并以此为基准给出参数Copula选择的小波方法.这得到以标普500指数、日经225指数、恒生指数和上证指数为样本的实证支持.进而从不同的时间尺度视角捕捉到股市之间潜在的相依模式.how to choose a copula is a key to study dependent structure of financial assets. Considering the local difference of dependency structure and the adaptive ability of wavelet function, soft threshold rules are introduced into copulas theory. A copulas estimator of soft threshold is put forward. Wavelet method is given to optimize para kkei 225 index, Hang metric copulas. It is supported by the empirical analysis on the S＆P 500 index, Niseng index and Shanghai index. It is confirmed the wavelet estimator be able to capture the market potential dependent models from a perspective of different time-scales.

关 键 词：小波分析 软阈值 参数Copula 非参数估计 

分 类 号：F830.91[经济管理—金融学]