基于贝叶斯推断的证券投资基金绩效分析

发布时间：2018-06-16 01:50

本文选题：证券投资基金 + 贝叶斯推断　；参考：《管理学报》2012年07期

【摘要】：从投资者先验信念的角度,运用贝叶斯推断对我国封闭式基金绩效进行实证研究。通过直观的问题引出投资者对基金管理者技巧的先验信念集合,并与具有4个风险指标的线性模型相结合,有效联系了投资决策过程中的重要因素——个人直观理念与实际市场数据,给出投资者在不同先验信念下的后验绩效。研究表明,随着投资者对管理者先验信念的增加,基金后验绩效增加,投资者更倾向于投资;只有当投资者对管理者有极强的先验信念时,才会投资于表现一般的基金;同理,不投资于绩优基金也需极强的先验信念。同时发现,基金费用对基金后验绩效的影响主要反映在基准线的平移。进一步,通过分时间段的投资组合权重分析,给出投资者在不同市场环境下的最优投资组合策略选择方法。
[Abstract]:From the perspective of investors' prior beliefs, this paper makes an empirical study on the performance of closed-end funds in China by using Bayesian inference. Through intuitionistic questions, it leads to the priori belief set of investor's skill of fund manager, and combines with the linear model with four risk indexes. This paper effectively links the important factors in the process of investment decision-personal intuitionistic idea and actual market data, and gives the posteriori performance of investors under different transcendental beliefs. Studies have shown that as investors' belief in managers' priori increases, fund posteriori performance increases, and investors are more likely to invest; only when investors have a strong priori belief in managers do they invest in funds that perform generally. Not investing in merit funds also requires a strong priori belief. At the same time, it is found that the impact of fund cost on fund posteriori performance is mainly reflected in the translation of the baseline. Furthermore, through the analysis of portfolio weight in different time periods, an optimal portfolio strategy selection method for investors in different market environments is presented.
【作者单位】：南京大学工程管理学院;
【基金】：国家自然科学基金资助重点项目(70932003);国家自然科学基金资助项目(70701016,71171109,71173098)
【分类号】：F224;F832.51

【参考文献】