运筹与管理2024，Vol.33Issue(12)：100-107,8.DOI:10.12005/orms.2024.0392
基于Smart Beta策略的PGP-MVS动态择时组合与绩效评价
PGP-MVS Dynamic Timing Portfolios and Performance Evaluation Based on Smart Beta
摘要
Abstract
After the financial crisis in 2008,investors preferred investment products with high transparency,sim-ple principles,low cost,strong personalization,and effective risk control.Therefore,Smart Beta strategies based on certain algorithms or rules emerge quickly and are widely accepted by the market.Relatively,the num-ber of Smart Beta strategic ETF funds in China is small,the size is small,the establishment time is short,and the strategy is simple.Based on the current situation and investors'preference for skewness in income distribu-tion,this article first establishes simulated ETF portfolios of four Smart Beta strategies:fundamental weighted portfolios(FW),equally weighted portfolios(EW),diversified minimum variance portfolios(DMV),and diversified risk parity portfolios(DRP).Based on these,a multiorder moments combination model is construc-ted,and then a polynomial goal programming model is used to select the optimal portfolios to explore the time-varying nature of different strategies.It reflects the actual impact of changes in the macroeconomic environment on portfolios returns,and is beneficial for investors to conduct risk management.Because this article extends the PGP method to Smart Beta portfolios,explores market timing styles for different Smart Beta portfolios,and obtains revenue paths through mechanisms conversion among mean,covariance and coskewness,it is defined as the PGP-MVS model. In data analysis section,first of all,this paper constructs simulated ETF portfolios of monthly rebalanced Smart Beta strategies using the CSI300 Index component stocks from May 2005 to June 2022 as samples,analy-zes the factor exposure and out-of-sample performance of these four simulated ETF portfolios,compares them with the market capitalization weighted CSI300 Index,and finds that the SP ratios of the four Smart Beta strategic portfolios are superior to the market capitalization weighted portfolios(CW).Under the Fama-French-Charhart four-factor model,the out-of-sample returns obtain positive alpha excess returns,which are statistically signifi-cant except FW,and are all greater than the alpha of CW quantitatively,as well as the daily risk-free interest rate of 0.000041.Secondly,in order to further explore the impact of different moment preference on portfolios selection,the PGP-MVS model with investor moment preference is used to conduct time varying dynamic portfolios selection for these five indexes and analyze their factor exposure,return paths,and out-of-sample performance under different moment preference,going beyond the traditional mean variance model framework.We use improved constant correlation estimation to estimate the covariance and coskewness of portfolios returns,which can reduce estimation parameters,improve portfolios returns out-of-sample,and overcome the curse of dimensionality of the traditional higher order moment estimation. The empirical analysis results show that the model is helpful in switching styles and creating revenue paths based on market risk changes.For example,when λ1=1,λ2=2and λ3=3,a very surprising phenomenon is found based on the weight distribution of the PGP-MVS portfolios.In the first five months of 2015,the weights of the PGP-MVS strategic portfolios were converted from the FW and DMV portfolios to the EW and DRP portfolios,which coincided with the actual situation of the undifferentiated limit up for 1000 stocks driven by leveraged funds at that time.The weights were converted to DMV and DRP portfolios from June to December of 2015 and the first four months of 2016,indicating that the PGP-MVS strategic portfolios were converted to risk-based port-folios during periods of severe volatility.During the 2018 Trade War,the weights of PGP-MVS strategic portfolios were mainly FW portfolios.Affected by the COVID-19 in the first three months of 2020,the weights of PGP-MVS strategic portfolios were mainly the combination of CW,DMV and DRP.This also indicated that after the market was impacted by the epidemic,the weights of the PGP-MVS strategic portfolios tilted towards diversified risk investment.In short,building Smart Beta strategic portfolios based on certain features and PGP-MVS strategic portfolios can not only enrich underlying instruments of investors with different risk preference,but also help investors reasonably anticipate and manage capital market risk,which is conducive to the healthy development of the capital market.关键词
Smart Beta/多项式目标规划/高阶矩/收益路径Key words
Smart Beta/polynomial goal programming/higher-order comoments/income paths分类
管理科学引用本文复制引用
周静..基于Smart Beta策略的PGP-MVS动态择时组合与绩效评价[J].运筹与管理,2024,33(12):100-107,8.基金项目
广东省哲学社会科学规划项目(GD20CYJ16) （GD20CYJ16）
