基于M-Copula-EGARCH-M-GED模型的相关风险度量及投资组合优化  

Measurement of Correlation Risk and Optimization of Portfolio Selection Based on M-Copula-EGARCH-M-GED Model

作　　者：宗钦原 申建平[1] 

机构地区：[1]重庆大学数学与统计学院,重庆401331

出　　处：《重庆理工大学学报（社会科学）》2015年第2期37-46,共10页Journal of Chongqing University of Technology（Social Science）

摘　　要：采用EGARCH-M模型进行单个资产建模,利用M-Copula函数构建资产的联合分布,对基于GED分布的联合资产收益率构建M-Copula-EGARCH-M模型,并采用Monte Carlo模拟方法计算出不同投资比例和置信水平的组合风险VaR和CVaR的值,并求出不同期望收益和置信水平下的最优组合投资权重。结果显示:当分位数为5%时,欲获得固定收益,同时风险最小,需将绝大部分资金投入深圳证劵市场;在分位数取10%时,目标收益率较低时,应将大部分资金投放在上海证劵市场。目标收益率较高时,上证市场的投资比例应降低,深证市场的资金比例应上升,具体的最优投资比例在实证部分已给出。实证结果表明:本文模型有利于投资者对投资权重的选择。We adopt EGARCH- M model for building a single asset model and used the M- Copula function to construct the joint distribution of assets, simultaneously, we built the M- CopulaEGARCH- M model for the return on joint assets which was based on GED distribution,then worked out the values of VaR and CVaR of portfolio risk under different investment ratios and different confidence levels by using Monte Carlo simulation method. And last we found the weights of the optimal investment portfolio under different expected revenue and confidence levels. To achieve the fixed income,while the minimal risk,the investors must put the vast majority of funds into the Shenzhen Securities Market when the quintile is 5 percent. When quintile is 10 percent,most of their moneyshould be put into Shanghai securities market if a target yield is low,otherwise the proportion of funds in Shenzhen Securities Market might increase and Shanghai market might reduce. In the empirical part,we gave a specific optimal investment ratio. The empirical results show that the proposed model is beneficial to investors when selecting investment weights.

关 键 词：M-Copula EGARCH-M Monte Carlo 投资组合 CVAR 

分 类 号：F224[经济管理—国民经济] F830