论文部分内容阅读
通过对常替代弹性资本资产定价模型中投资标度问题的分析,提出了Copula贝叶斯估计方法用以获得系统风险β与投资标度比λ的联合后验分布.Copula贝叶斯估计方法针对数据非正态特征及强相关性特征而构建,采用Copula函数取代原有普通贝叶斯估计方法中的正态假设.传统贝叶斯估计方法假设了正态的似然函数,忽略了数据可能存在尖峰后尾等在金融实证数据分析中普遍存在的非正态情况.Copula贝叶斯估计算法采用半相依回归法处理数据的强相关性问题,将原有函数依照数据形式假设为非正态结构.针对来自6个工业产业24组公司数据的系统风险参数β与其对应的投资标度参数比λ进行估计,获得不同行业中系统风险参数与投资标度之间的动态关系并进行分析,为业界投资及相关研究提供有效参考建议.
Through the analysis of the investment scale problem in the constant alternative elastic capital asset pricing model, a Copula Bayesian estimation method is proposed to obtain the joint posterior distribution of the systematic risk β and the investment scale ratio λ. Data non-normal features and strong correlation features to build the use of Copula function to replace the original normal Bayesian estimation method in the normal hypothesis.Traditional Bayesian estimation method assumes a normal likelihood function, ignoring the data may be There is a non-normal situation such as spikes and back-end in the financial empirical data analysis.Copula Bayesian estimation algorithm using semi-dependent regression method to deal with the strong correlation of data, the original function according to the data form is assumed to be non-normal Structure.According to the system risk parameter β and its corresponding investment scale parameter ratio λ from the data of 24 groups of six industrial industries, the dynamic relationship between the system risk parameter and investment scale in different industries is obtained and analyzed, which is Industry investment and related research to provide effective reference suggestions.