基于分形和S型效用的选股策略及M-CVaR最优资产配置
Stock Selection Strategy Based on Fractal and S-Type Utility and M-CVaR Optimal Asset Allocation
孙景云 1马小雯2
作者信息
- 1. 兰州财经大学统计与数据科学学院,兰州 730020;甘肃经济发展数量分析研究中心,兰州 730020
- 2. 兰州财经大学统计与数据科学学院,兰州 730020
- 折叠
摘要
以上证50指数的主要成分股为研究对象,首先基于分形与S型效用理论构建股票风险综合评估指标,作为构建投资组合的选股依据.然后在金融资产收益率服从非对称拉普拉斯分布的假设下,采用CVaR值度量投资组合的风险,进而构建M-CVaR最优资产配置模型,并将该模型转化为二次规划问题进行求解.在实证分析阶段,利用滑动窗口法分别以月、季度、半年和一年为周期对最佳股票投资集的最优配置比例进行动态调整.结果表明,利用分形与S型期望效用理论筛选出的部分股票投资集可以获得比全部股票投资集更优的投资收益,且发现调整周期为1年的资产配置方案能获得较其他调整周期更高的累计收益率和夏普比率.
Abstract
Taking the main components of the SSE 50 Index as the research ob-ject,this paper first constructs a comprehensive evaluation index of stock risk based on fractal and S-utility theory,which is the basis of selecting stocks for portfolio construction.Then,under the assumption that the return on financial assets obeys asymmetric Laplacian distribution,the CVaR value is adopted to measure the risk of portfolio,and then an M-CVaR optimal asset allocation model is constructed,which is transformed into a quadratic programming problem to solve.In the empirical analysis stage,the sliding window method is used to dynamically adjust the optimal allocation ratio of the best stock set with monthly,quarterly,semi-annual and one-year cycles respectively.The results show that some stock investment sets screened by fractal and S-type expected utility theory can obtain better investment returns than all stocks participating in the portfolio,and the asset allocation scheme with one-year adjustment cycle can obtain higher cumulative return and Sharpe ratio than other adjustment cycles.
关键词
分形理论/S型期望效用/非对称拉普拉斯分布/CVaRKey words
Fractal theory/S-type expected utility/asymmetric Laplacian distribu-tion/CVaR引用本文复制引用
基金项目
国家自然科学基金(72061020)
甘肃省科技计划(21JR1RA280)
陇原青年创新创业人才项目(2022)()
出版年
2024
NETL
NSTL
万方数据