Multistage Portfolio Optimization Based on Several Risk Measures

doi:10.12005/orms.2017.0257

Abstract

Abstract: This paper utilizes Mean-Variance(MV), VaR(Value at Risk), CVaR(Conditional VaR), HMCR(p=1,2,3)(Higher Moment Coherent Risk)to build multistage portfolio optimization. Firstly, testing the ability of discrimination for risk from the perspective of the coherent axiom and the consistence with stochastic dominance, and the HMCR(p=2,3) are much better than other risk measures. Then, we introduce the risk averse planning function under static and dynamic condition, and conditional versions of CVaR and HMCR. Finally, we give an empirical analyzes of constituent stocks of SSE 50 index and the index itself, and according to the cumulative returns of these risk measures portfolios and SSE 50 index, we find that multistage portfolios are better than single-stage portfolios, HMCR(p=2,3) are much better than SSE 50 index and the other risk measures. In addition, we consider the investors’ investment decision, and find that the positive investment strategy of HMCR(p=2,3) can get a high cumulative returns in the stable period, peak period and decline period of stock market, and the negative investment of index would be much better in the rise period of stock market.

Key words: risk measure, coherent axiom, consistence of stochastic dominance, multistage portfolio optimization, investment decision

摘要： 基于均值-方差(MV)、VaR(Value at Risk)、CVaR(Conditional VaR)、HMCR(p=1,2,3)(Higher Moment Coherent Risk)几种风险测度进行多阶段组合优化研究。首先从一致性公理和随机占优一致性角度分析几种风险测度的风险识别能力,认为HMCR(p=2,3)的风险识别能力最高,然后给出静态和动态下的风险规避型的规划函数及多阶段CVaR和HMCR模型,最后依据单阶段和多阶段优化模型,对上证50指数成份股进行实证分析。对比单阶段和多阶段下几种风险测度优化组合的累计收益率及几种风险测度之间的关系,结合上证50指数收益率发现,多阶段优化组合要整体优于单阶段优化组合,且HMCR(p=2,3)要优于指数收益率和其它几种风险测度。从投资者投资决策方面来分析,HMCR(p=2,3)型积极投资策略比较适用于股市平稳期、顶峰期和下降期,被动投资策略比较适用于股市上升期。

关键词: 风险测度, 一致性公理, 随机占优一致, 多阶段组合优化, 投资策略

CLC Number:

F830

ZHENG Cheng-li, YAO Yin-hong. Multistage Portfolio Optimization Based on Several Risk Measures[J]. Operations Research and Management Science, 2017, 26(11): 35-41.

郑承利, 姚银红. 基于几种风险测度的多阶段组合优化研究[J]. 运筹与管理, 2017, 26(11): 35-41.

References

[1] Markowitz H. Portfolio selection[J]. The journal of Finance, 1952, 7(1): 77-91.
[2] Jorion P. Value at risk[M]. 2nd Edition. 2001, New York: McGraw-Hill.
[3] Duffie D, Pan J. An Overview of value at risk[J]. Journal of Derivatives, 1997: 7- 49.
[4] Artzner P, Delbaen F, Eber J M, Heath D. Coherent measures of risk[J]. Mathematical Finance, 1999, 9(3): 203-228.
[5] Rockafellar R T, Uryasev S. Conditional value-at-risk for general distributions[J]. Journal of Banking & Finance, 2002, (26): 1443-1471.
[6] Kusuoka S. On law invariant coherent risk measures[J]. Advances in Mathematical Economics, 2001, 3(1215): 83-95.
[7] Wong W K, Ma C. Preferences over meyer’s location-scale family[J]. Economic Theory, 2008, 37(1): 119-146.
[8] Wong W K, Ma C. Stochastic dominance and risk measure[J]. A decision-theoretic foundation for VaR and C-VaR[J]. European Journal of Operational Research, 2010, 207(2): 927-935.
[9] Föllmer H, Schied A. Convex measures of risk and trading constraints[J]. Finance and Stochastics, 2002, 6(4): 429- 447 .
[10] Föllmer H, Schied A. Stochastics finance: an introduction in discrete time[M]. 2nd Ed., 2004, Walter de Gruyter.
[11] Föllmer H, Schied A. Convex and coherent risk measures[J]. preprint, Humboldt University(2008).
[12] Frittelli M, Gianin E R. Putting order in risk measures[J]. Journal of Banking and Finance, 2002, 26(7): 1473-1486.
[13] 郑文通.金融风险管理的VAR方法及其应用[J].国际金融研究,1997,(9):58- 62.
[14] 戴国强,徐龙炳,陆蓉.VaR方法对我国金融风险管理的借鉴及应用[J].金融研究,2000,(7):45-51.
[15] 刘小茂,李楚霖,王建华.风险资产组合的均值-CVaR有效前沿(I)[J].管理工程学报,2003,17(1):29-33.
[16] 刘小茂,李楚霖,王建华.风险资产组合的均值-CVaR有效前沿(II)[J].管理工程学报,2005,19(1):1-5.
[17] 田新民,黄海平.基于条件VaR(CVaR)的投资组合优化模型及比较研究[J].数学的实践与认识,2004,34(7):39- 49.
[18] 刘俊山.基于风险测度理论的VaR与CVaR的比较研究[J].数量经济技术经济研究,2007,24(3):125-133.
[19] 唐爱国,秦宛顺.广义随机占优单调一致风险测度和ES(n)[J].金融研究,2003,(4):84-93.
[20] Krokhmal P A. Higher moment coherent risk measures[J]. Quantitative Finance, 2007, (7): 373-387.
[21] Chen Z P, Wang Y. A new class of coherent risk measures based on p-norms and their applications[J]. Appl. Stochastic Models Bus. Ind., 2007, (23): 49- 62.
[22] Dentcheva D, Penev S, Ruszczyński A. Kusuoka representation of higher order dual risk measures[J]. Annals of Operations Research, 2010, (181): 325-335.
[23] 郑承利,姚银红.基于高阶一致风险测度的组合优化[J].系统管理学报,2017,26(05):857- 868.
[24] Riedel F. Dynamic coherent risk measures[J]. Stochastic Process and their Applications, 2004, 112(2): 185-200.
[25] Föllmer H, Knispel T. Convex risk measures: basic facts, law-invariance and beyond, asymptotics for large portfolios[J]. World Scientific, Chapter Part II, 2013: 507-554.
[26] Krokhmal P, Zabarankin M, Uryasev S. Modeling and optimization of risk[J]. Surveys in Operations Research and Management Science, 2011, 16(2): 49- 66.
[27] Artzner P, Delbaen F, Eber J M, Heath D, Ku H. Coherent multiperiod risk measurement[J]. Preprint, 2002.
[28] Artzner P, Delbaen F, Eber J M, Heath D, Ku H. Coherent multiperiod risk adjusted values and Bellman’s principle[J]. Annuals of Operations Research, 2007, 152(1): 5-22.
[29] Matmoura Y, Penev S. Multistage optimization of option portfolio using higher order coherent risk measures[J]. European Journal of Operational Research, 2013, (227): 190-198.
[30] Shapiro A, Dentcheva D, Ruszczyński A. Lectures on stochastic programming: modelling and theory[M]. SIAM. 2009.
[31] Bodie Z, Kane A, Marcus A J.投资学[M].汪昌云,张永冀等,译.9版.北京:机械工业出版社,2012.