| 640 | 10 | 94 |
| 下载次数 | 被引频次 | 阅读次数 |
在高阶矩投资组合中,使用传统样本估计方法会产生较高估计误差和模型设定误差。本文在多因素模型的基础上,给出一种改进的协高阶矩估计方法,分析了基于多因素模型压缩估计量的渐进一致性。蒙特卡洛模拟表明,多因素压缩估计量在有限样本中具有更小的平均绝对误差、根均方误差以及更高的平均绝对改进百分比,有效提高了协高阶矩矩阵估计的精度;即使在样本观测量比资产数目少时,估计的协高阶矩矩阵精度都会有较大提高。基于2005年6月至2019年5月沪深300成分股的高阶矩投资组合实证发现,多因素压缩方法与其他估计方法相比,在年化收益率上可以获得4.7%～32.8%的提升,最大回撤能够下降3.7%～18.3%,表明使用多因素压缩估计方法构建的投资组合有更大的可能获得更多货币效用增益,以及面临亏损时,产生的最大亏损更小。该方法有助于金融机构或理性投资者在进行投资组合时减小投资损失,获得更好的投资回报。
Abstract：In the higher-order moment portfolio, the traditional estimation method will have larger estimation error and model setting error. Based on the multi-factor model,this paper gives an improved estimation method for higher-order comoment matrices and the asymptotic consistency of the shrinkage estimators. The Monte Carlo simulation shows that the multi-factor shrinkage estimator has smaller average absolute error,root mean square error,and a higher percentage improvement in average absolute error in a finite sample. It effectively improves the accuracy of the higher-order comoment matrix estimation,even when the sample observation is smaller than the number of assets. Based on the empirical evidence of higher-order moment portfolios of CSI 300 stocks from June 2005 to May 2019,compared with other estimation methods,the multi-factor shrinkage method can improve 4. 7% ～ 32. 8% in annualized return and the maximum drawdown can fall by 3. 7% ～ 18. 3%. It indicates that the investment portfolio constructed by the multi-factor shrinkage estimation method has a higher possibility to obtain a greater monetary utility gain,and the maximum loss,if in a loss,is smaller. This method helps financial institutions or rational investors to reduce investment losses and obtain better investment returns when investing in portfolios.
[1]贾婧，鲁万波，柯睿．基于回归的时变偏度和时变峰度识别检验[J]．统计研究，2018(11):116-128．
[2]蒋翠霞，许启发，张世英．金融市场条件高阶矩风险与动态投资组合[J]．中国管理科学，2007(1):27-33．
[3]李志冰，杨光艺，冯永昌，等．Fama-French五因素模型在中国股票市场的实证检验[J]．金融研究，2017(6):191-206．
[4]彭胜志．基于高阶矩的投资组合优化研究[M]．第一版．北京:中国林业出版社，2016．
[5]秦磊，郁静，孙强．混频时间序列的潜在因素分析及其应用[J]．统计研究，2019(9):104-114．
[6]Azzalini A,Capitanio A.Distributions Generated by Perturbation of Symmetry with Emphasis on a Multivariate Skew T-distribution[J].Journal of the Royal Statistical Society Series B,2003,65(2):367-389．
[7]Boudt K,Cornilly D,Verdonck T.A Coskewness Shrinkage Approach for Estimating the Skewness of Linear Combinations of Random Variables[J].Journal of Financial Econometric,2018,10:1-23．
[8]Boudt K,Lu W B,Peeters B.Higher Order Comoments of Multifactor Models and Asset Allocation[J].Finance Research Letters,2015,13(5):225-233．
[9]Boudt K,Peterson B,Christophe C.Estimation and Decomposition of Downside Risk for Portfolios with Non-normal Returns[J].Journal of Risk,2008,11(2):79-103．
[10]Fama E F,French K R.A Five-factor Asset Pricing Model[J].Journal of Financial Economics,2015,116(1):1-22．
[11]Fan J Q,Fan Y Y,Lv J C.High Dimensional Covariance Matrix Estimation Using a Factor Model[J].Journal of Econometrics,2008,147(1):186-197．
[12]Harvey C R,et al.Portfolio Selection with Higher Moments[J].Quantitative Finance,2010,10(5):469-485．
[13]Jagannathan R,Ma T.Risk Reduction in Large Portfolios:Why Imposing the Wrong Constraints Helps[J].Journal of Finance,2003,58(4):1651-1684．
[14]Jondeau E,Rockinger M.Optimal Portfolio Allocation under Higher Moments[J].European Financial Management,2006,12(1):29-55．
[15]Jondeau E,Rockinger M.Conditional Volatility,Skewness,and Kurtosis:Existence,Persistence,and Comovements[J].Journal of Economic Dynamics and Control,2003,27(10):1699-1737．
[16]Kim T,White H.On More Robust Estimation of Skewness and Kurtosis[J].Finance Research Letters,2004,1(1):56-73．
[17]Lambert P,Laurent S.Modelling Financial Time Series Using GARCH-type Models and a Skewed Student Density[R].Stat Discussion,Universite Catholique de Louvain,Belgium,2001．
[18]Ledoit O,Wolf M.Improved Estimation of the Covariance Matrix of Stock Returns with an Application to Portfolio Selection[J].Journal of Empirical Finance,2003,10(5):603-621．
[19]Ledoit O,Wolf M.Honey,I Shrunk the Sample Covariance Matrix[J].Journal of Portfolio Management,2004,30(4):110-119．
[20]Markowitz H.Portfolio Selection[J].Journal of Finance,1952,7(1):77-91．
[21]Martellini L,Ziemann V.Improved Estimates of Higher-order Comoments and Implications for Portfolio Selection[J].Review of Financial Studies,2010,23(4):1467-1502．
[22]Mooijaart A.Factor Analysis for Non-normal Variables[J].Psychometrika,1985,50(3):323-342．
[23]Ross S A.The Arbitrage Theory of Capital Asset Pricing[J].Journal of Economic Theory,1976,13(3):341-360．
[24]Sun Q,Yan Y X.Skewness Persistence with Optimal Portfolio Selection[J].Journal of Banking and Finance,2003,27(6):1111-1121．
[25]William H G.Econometric Analysis[M].6th edn.Pearson Prentice Hall,New Jersey,2010.
(1)因篇幅所限，该处各种估计方法所需估计参数个数比较以附表1展示，见《统计研究》网站所列附件。下同。
(1)本文对所有股票基于五因素模型进行回归，并获得不同股票的因素载荷，通过将因素载荷之和由大到小排序，选择前15只具有最大因素载荷的股票作为参数校准的股票。15只股票分别为:金地集团，中联重科，长春高新，华域汽车，国投电力，江西铜业，河钢股份，方大炭素，海螺水泥，西南证券，金融街，五粮液，华东医药，广汇汽车，圆通速递。
(2)因篇幅所限，具体估计值以附表2展示。
(1)因篇幅所限，以附表3~5展示。
(2)因篇幅所限，以附表6~8展示。
(1)因篇幅所限，以附图1展示。
(2)因篇幅所限，描述性统计以附表9展示。
基本信息:
DOI：10.19343/j.cnki.11-1302/c.2020.12.008
中图分类号:F224;F832.51
引用信息:
[1]鲁万波,王建业.基于因素模型的协高阶矩矩阵估计及其应用研究[J].统计研究,2020,37(12):105-121.DOI:10.19343/j.cnki.11-1302/c.2020.12.008.
基金信息:
国家自然科学基金面上项目“高维协高阶矩的估计及其在投资组合中的应用”(71771187);国家自然科学基金国际合作交流项目“基于高频、混频数据的高阶矩投资组合研究”(72011530149);; 教育部“新世纪优秀人才支持计划”(NCET-13-0961);; 中央高校基本科研业务费专项资金“金融市场计量经济前沿研究”(JBK190602)
2020-12-31
2020-12-31
2020-12-31