基于Frank-Copula贝叶斯估计的系统性风险对样本标度的敏感性分析

研究表明数据样本标度的选取对系统性风险估计结果具有直接影响，本文基于150只A股样本数据构建25组投资组合，通过Copula贝叶斯估计方法获得系统风险β与投资标度比λ无信息先验的联合后验分布，对系统风险β与投资标度比λ的影响效应进行了分析。研究表明，样本标度对我国A股市场的系统性风险值估计存在误差性影响，且该影响随着公司规模的增加而不断上升，即样本标度的选取对于大额机构投资者影响较大，但数据对于账面市值比不敏感。以月度数据为基础数据，我国市场系统性风险值与样本标度比存在弱负相关关系，相比较以机构投资者占据资本市场主流的美国数据，我国市场真实投资标度存在明显差异。

Investment Horizon,System Risk Value and the Sensitive Effect

The effects of investment horizon on the estimate of the systematic risk are first investigated by Brennan et al,(2012). In general, the true investment horizon is unknown. The empirical work will overestimate the coefficient of the systematic risk based on the observed horizon. The copula Bayesian estimation approach is proposed to get the posterior distribution of the coefficients of the system risk beta and the investment horizon ratio gama in the Fama-French three factor model. The potent problem of the traditional Bayesian estimation is that the assumption of normal likelihood function ignores some fluctuations such as high peak and fat tail relative to kurtosis and skewness, which have been frequently, reported in financial data analyses. The copula Bayesian approach instead of the traditional Bayesian estimation is built to consider the pattern of the data with the strong correlation and the non-normal distributions. The reason why the copula function is chosen is to fit the pattern of the data. In the empirical work,the interaction of the system risk and the investment horizon is analyzed in 25 portfolios from 150 different data. Compared with the U.S. data, the correlation of the systemic risk and investment horizon is negative, and the frequency of the true horizon is higher than observed one in China. With the increase of the size of the company, the effect of the investment horizon is obviously magnified. And the appearance leads to the estimation bias of the systemic risk.