基于指数跟踪的投资组合优化模型及实证分析
发布时间：2019-06-12 18:43
【摘要】：指数化投资是被动投资管理领域中最重要的投资策略之一,广泛用于指数基金及其它机构投资者的资产管理.作为指数化投资的具体管理形式,指数跟踪技术得到了现代投资组合理论的有力支持,越来越多地引起了学术界和金融业界的重视. 选择合适的复制方法是指数跟踪的核心问题.本文结合了因素模型和均值-方差模型的思想,提出了带基数约束的多因素指数跟踪模型.我们利用多因素模型来刻画股票的收益,以市场组合的贝塔值为基准,对跟踪组合贝塔值进行控制的同时最小化投资组合的风险,得到两种多因素指数跟踪模型.考虑到实际投资需要,我们在模型中引入跟踪组合规模约束和持有量约束,所得模型等价于混合0-1二次规划问题.为了更快求解该模型,我们从原问题的拉格朗日对偶出发,利用锥优化理论获得一类改进的混合0-1二次规划等价模型.数值试验表明,新等价模型在计算效率方面具有较明显的优势.为说明多因素跟踪模型的实际表现,我们利用中国沪深300指数的真实数据构建了3个目标指数,并将该模型与均值-方差指数跟踪模型进行了比较分析.基于沪深300指数的实证跟踪结果表明,我们的指数跟踪模型能够产生较优的跟踪组合,并且具有很好的准确性和灵活性.最后,我们综合考虑了跟踪投资组合的绝对风险和相对风险,对多因素指数跟踪模型进行进一步推广,并对推广后的模型进行实证分析. 本文总共分为六章,第一章介绍了被动投资管理与指数化投资方法.第二章介绍了指数跟踪的研究成果和基本的模型,如均值-方差指数跟踪模型,均值-绝对偏差指数跟踪型和单因素指数跟踪模型等.第三章重点介绍了带基数约束的多因素指数跟踪模型及其求解方法.在第四章中,我们基于中国证券市场的历史数据对多因素指数跟踪模型进行实证分析.在第五章中,我们将探讨推广后的多因素指数跟踪模型并进行实证分析.第六章是结论部分,是对本文内容的总结以及对未来研究的展望.
[Abstract]:Indexed investment is one of the most important investment strategies in the field of passive investment management, which is widely used in the asset management of index funds and other institutional investors. As the specific management form of indexed investment, index tracking technology has been strongly supported by modern portfolio theory, and has attracted more and more attention from academia and financial circles. Choosing the appropriate replication method is the core problem of exponential tracking. Based on the idea of factor model and mean-variance model, a multi-factor exponential tracking model with cardinality constraint is proposed in this paper. We use the multi-factor model to depict the return of the stock. Based on the beta value of the market portfolio, we control the beta value of the tracking portfolio while minimizing the risk of the portfolio, and obtain two kinds of multi-factor index tracking models. Considering the actual investment needs, we introduce the tracking portfolio size constraint and holding constraint into the model, and the obtained model is equivalent to the mixed 0 鈮，
本文编号：2498221
[Abstract]:Indexed investment is one of the most important investment strategies in the field of passive investment management, which is widely used in the asset management of index funds and other institutional investors. As the specific management form of indexed investment, index tracking technology has been strongly supported by modern portfolio theory, and has attracted more and more attention from academia and financial circles. Choosing the appropriate replication method is the core problem of exponential tracking. Based on the idea of factor model and mean-variance model, a multi-factor exponential tracking model with cardinality constraint is proposed in this paper. We use the multi-factor model to depict the return of the stock. Based on the beta value of the market portfolio, we control the beta value of the tracking portfolio while minimizing the risk of the portfolio, and obtain two kinds of multi-factor index tracking models. Considering the actual investment needs, we introduce the tracking portfolio size constraint and holding constraint into the model, and the obtained model is equivalent to the mixed 0 鈮，
本文编号：2498221
本文链接：https://www.wllwen.com/guanlilunwen/huobilw/2498221.html
