«上一篇/Previous Article|本期目录/Table of Contents|下一篇/Next Article»

《金融教育研究》[ISSN:1006-6977/CN:61-1281/TN]

卷:

期数:

2013年06期

页码:

34-36

栏目:

出版日期:

2013-12-31

文章信息/Info

Title:

Bivariate Copula Function of Portfolio Model Selection

作者:

张学仁;

江西财经大学信息管理学院,江西南昌,330032

Author(s):

ZHANG Xueren

关键词:

投资组合; 二元正态Copula; 二元t-Copula; 距离比较

Keywords:

investment portfolio;  bivariate normal Copula;  bivariate t-Copula;  distance comparison

分类号:

F830.9

文献标志码:

A

摘要:

Coplua模型是组合投资风险评估中常用模型,它具有多种不同的类型,模型选择的好坏对风险评估结果具有至关重要的影响。本文主要比较了二元正态Copula模型和二元t-Copula模型对中国股市数据拟合的优劣程度。针对这两种模型,利用上证综指、深证成指、上证基金、深证基金、东风汽车、中国石化、宝钢股份和万家乐的日收盘价数据估计相应的参数得到相应的拟合分布,然后分别与经验Copula函数作比较,通过计算拟合分布与经验分布之间的距离,得出二元t-Copula函数能更好地拟合两组投资组合的日收益率数据的结论。

Abstract:

Coplua model is commonly used in the portfolio risk assessment,it has many different types of models,stand or fall of model selection has a crucial influence on the risk assessment result.This essay compares bivariate normal Copula model with bivariate t-Copula model on China's strengths and weaknesses of the stock market by fitting the data.For both models,using Shanghai Composite Index,Shenzhen Component Index,SSE Fund,Shenzhen Securities Fund,Dongfeng Motor,Sinopec,Baosteel and Macro's closing price data estimates the corresponding parameters of the corresponding fitting distribution,then compares with the experience of Copula functions separately.By calculating the distance between fitted distribution and the empirical distribution,the essay comes to a conclusion that the bivariate t-Copula function is better fitting the daily yield rate data of the two investment portfolios.

常用功能

更新日期/Last Update: 1900-01-01