一种增强型指数追踪模型设计及应用

发布时间：2018-05-07 10:03

  本文选题：指数追踪 + 折中路径　； 参考：《数量经济技术经济研究》2017年05期

【摘要】：研究目标:构建了可以调节追踪误差和超额收益的增强型指数追踪模型,并给出了广义最小角度回归算法(GLARS),用以计算调节参数作用下模型解的折中路径。研究方法:通过模拟数据和五组世界主要股票市场指数的历史数据,对本文提出的模型和算法与同类模型和算法进行了性能比较;同时追踪上证50指数构建若干稀疏且稳定的资产组合模型,通过信息比率等指标对投资组合进行评价。研究发现:本文构建的模型可用以构造权衡追踪效果和超额收益,且稀疏的资产组合,GLARS算法相对传统预设参数的算法具有良好的求解能力和计算速度。研究创新:引入调节参数平衡追踪效果和超额收益,并针对中国股票市场的特点,在增强型指数追踪模型施加非负约束;GLARS算法可遍历所有折中意义下的最优解。研究价值:本文提出的增强型指数追踪模型在国内具有较强适用性,在保证资产稀疏性的前提下可以得到超额收益,同时丰富了目前投资组合中的方法论研究。
[Abstract]:Research objective: an enhanced exponential tracking model which can adjust the tracking error and excess return is constructed, and the generalized minimum angle regression algorithm is given to calculate the compromise path of the model under the action of adjusting parameters. Methods: through the simulation data and five groups of historical data of the world's major stock market indexes, the performance of the proposed model and algorithm is compared with that of the similar models and algorithms. At the same time, several sparse and stable portfolio models are constructed by tracking the Shanghai Stock Exchange 50 Index, and the portfolio is evaluated by information ratio and other indicators. It is found that the model constructed in this paper can be used to construct tradeoff between tracing effect and excess return, and the sparse portfolio GLARS algorithm has good solving ability and computing speed compared with the traditional algorithm with preset parameters. Research innovation: according to the characteristics of Chinese stock market and the characteristics of Chinese stock market, the GLARS algorithm can traverse the optimal solution in the sense of all compromises by introducing the adjusted parameter equilibrium tracking effect and excess return, and according to the characteristics of the Chinese stock market, applying non-negative constraints to the enhanced exponential tracking model. Research value: the enhanced index tracking model proposed in this paper has strong applicability in China, which can obtain excess returns on the premise of asset sparsity, and enriches the current research on portfolio methodology.
【作者单位】： 中央财经大学统计与数学学院;
【基金】：国家自然科学基金项目(71403310) 北京市社会科学基金项目(16LJB005) 中央财经大学青年科研创新团队支持计划;中央财经大学博士研究生重点选题支持计划的资助 中央高校基本科研业务经费
【分类号】：F224.0
，

本文编号：1856488