基于参数二次规划和逆向优化的超越股票市场指数的实证研究
详细信息 查看官网全文
- 英文论文题名：Directly Evaluating Genetic Algorithms for Portfolio Optimization in the Context of Big Data
- 论文作者：齐岳 ; 王治皓 ; 林龙
- 英文论文作者：QI Yue ; WANG Zhi-hao ; LIN Long ; China Academy of Corporate Governance ; Nankai University ; Department of Financial Management ; Business School ; Nankai University ; Department of Business Administration ; Fujian Commercial College
- 年：2016
- 作者机构：南开大学中国公司治理研究院;南开大学商学院;福建商学院工商管理系;
- 论文关键词：参数二次规划 ; 投资组合 ; 逆向优化
- 英文论文关键词：parametric quadratic programming ; inverse optimization ; portfolio
- 会议召开时间：2016-11-12
- 会议录名称：第十八届中国管理科学学术年会论文集
- 英文会议录名称：Proceedings of the 18th Chinese Annual Conference on Management Science
- 语种：中文
- 分类号：F224;F832.51
- 学会代码：ZHYJ
- 会议名称：第十八届中国管理科学学术年会
- 会议地点：中国陕西西安
- 主办单位：中国优选法统筹法与经济数学研究会、西安交通大学、中国科学院科技战略咨询研究院、《中国管理科学》编辑部
- 学会名称：中国优选法统筹法与经济数学研究会
- 页数：5
- 文件大小：499k
- 原文格式：O
- 会议级别：全国
摘要
习近平总书记在2015年底提到股票市场的发展,对股票市场的发展提出了新的要求,股票市场的健康发展需要投资者主动控制投资组合的风险,这促使投资组合在我国股票市场的应用更加广泛,更对投资组合模型的效果研究提出了新的要求。为了科学系统地探索投资组合模型在我国股票市场中的应用效果,本文选取了自2013年1月至2015年12月中由证监会划分的18个行业中流通市值最大的股票,以此建立带上界约束条件的投资组合模型。针对股票市场指数的收益率,本文创新性地使用参数二次规划和逆向优化计算出精确的权重向量的函数表达式,求解出特定收益率下最优投资组合的权重向量,进而对比在下一时间阶段内我们所构造的投资组合与股票市场指数收益率。12个实验组的分析结果表明,基于参数二次规划和逆向优化下的投资组合收益率基本超越了同时期的股票市场指数的收益率,实证表明投资组合模型在我国股票市场有着一定的应用价值。
President Xi Jinping put forward new requirement of development of stock market when he mentioned stock market development on the end of 2015.Investors control investment portfolio actively is important for developing stock market healthily,which makes investment portfolio to be widely used in stock market.At the same time,the significance of investment portfolio makes the research on portfolio model to achieve a higher level.To explore the effectiveness of applying portfolio model to Chinese stock market,this study selects the largest stock from 18 industries divided by CSRC from January 2013 to December2015,combining with portfolio model with upper bound constraint.The innovation of this study is to use parametric quadratic programming and inverse optimization to calculate the exact weight vector function in the optimal portfolio according to a specific expected return,and we compare the return of the portfolio and the stock market index.The 12 control experimental groups' result shows that based on parametric quadratic programming and inverse portfolio optimization,the return of portfolio exceeds the return of stock market index.This study proves that the portfolio model has a certain application value in Chinese stock market.
President Xi Jinping put forward new requirement of development of stock market when he mentioned stock market development on the end of 2015.Investors control investment portfolio actively is important for developing stock market healthily,which makes investment portfolio to be widely used in stock market.At the same time,the significance of investment portfolio makes the research on portfolio model to achieve a higher level.To explore the effectiveness of applying portfolio model to Chinese stock market,this study selects the largest stock from 18 industries divided by CSRC from January 2013 to December2015,combining with portfolio model with upper bound constraint.The innovation of this study is to use parametric quadratic programming and inverse optimization to calculate the exact weight vector function in the optimal portfolio according to a specific expected return,and we compare the return of the portfolio and the stock market index.The 12 control experimental groups' result shows that based on parametric quadratic programming and inverse portfolio optimization,the return of portfolio exceeds the return of stock market index.This study proves that the portfolio model has a certain application value in Chinese stock market.
引文
[1]Markowitz H.Portfolio Selection[J].Journal of Finance,1952,7(1):77-91.
[2]Markowitz H.The optimization of a quadratic function subject to linear constraints[J].Naval Research Logistics Quarterly,1956,3(1-2):111-133.
[3]Markowitz H,Todd G P.Mean-Variance analysis in portfolio choice and capital markets[M],Wiley,2000.
[4]Best M J.An algorithm for the solution of the parametric quadratic programming problem[J].Applied Mathematics&Parallel Computing,1996:57-76.
[5]Best M J,Kale J.Quadratic programming for largescale portfolio optimization[J].Financial Services Information Systems,CRC Press,2000,513-529.
[6]Stein M,Branke J,Schmeck H.Efficient implementation of an active set algorithm for large-scale portfolio selection[J].Computers&Operations Research,2008,35(12):3945-3961.
[7]Hirschberger M,Qi Y,Steuer R E.Large-scale MV efficient frontier computation via a procedure of parametric quadratic programming[J].European Journal of Operational Research,2010,204(3):581-588.
[8]Steuer R E,Qi Y,Hirschberger M.Comparative issues in large-scale mean-variance efficient frontier computation[J].Decision Support Systems,2011,51(2):250-255.
[9]程聪,张立卫.二次规划逆问题的牛顿方法[J].运筹学学报,2014,18(3):60-70.
[10]马建华.参数规划的逆问题[J].山东大学学报:理学版,2003,38(5):40-44.
[11]Engl H W,Ramlau R.Regularization of inverse problems[M].Dordrecht:Kluwer Academic Publisher Group,1996.
[12]Ahuja R K,Orlin J B.Inverse Optimization[J].Operations Research,2001,49(5):771-783.
[13]Zagst R,Poschik M.Inverse portfolio optimisation under constraints[J].Journal of Asset Management,2008,9(3):239-253.
[14]齐岳,林龙.基于社会责任的多目标证券投资组合模型[J].中国流通经济,2013,27(8):119-123.
[15]Steuer R E,Qi Yue,Hirschberger M.et al.Suitable portfolio investors,nondominated frontier sensitivity,and effects of multiple objectives on portfolio selection[J].Annals of Operations Research,Issue on Financial Optimization,2007,152(1):297-317.
[2]Markowitz H.The optimization of a quadratic function subject to linear constraints[J].Naval Research Logistics Quarterly,1956,3(1-2):111-133.
[3]Markowitz H,Todd G P.Mean-Variance analysis in portfolio choice and capital markets[M],Wiley,2000.
[4]Best M J.An algorithm for the solution of the parametric quadratic programming problem[J].Applied Mathematics&Parallel Computing,1996:57-76.
[5]Best M J,Kale J.Quadratic programming for largescale portfolio optimization[J].Financial Services Information Systems,CRC Press,2000,513-529.
[6]Stein M,Branke J,Schmeck H.Efficient implementation of an active set algorithm for large-scale portfolio selection[J].Computers&Operations Research,2008,35(12):3945-3961.
[7]Hirschberger M,Qi Y,Steuer R E.Large-scale MV efficient frontier computation via a procedure of parametric quadratic programming[J].European Journal of Operational Research,2010,204(3):581-588.
[8]Steuer R E,Qi Y,Hirschberger M.Comparative issues in large-scale mean-variance efficient frontier computation[J].Decision Support Systems,2011,51(2):250-255.
[9]程聪,张立卫.二次规划逆问题的牛顿方法[J].运筹学学报,2014,18(3):60-70.
[10]马建华.参数规划的逆问题[J].山东大学学报:理学版,2003,38(5):40-44.
[11]Engl H W,Ramlau R.Regularization of inverse problems[M].Dordrecht:Kluwer Academic Publisher Group,1996.
[12]Ahuja R K,Orlin J B.Inverse Optimization[J].Operations Research,2001,49(5):771-783.
[13]Zagst R,Poschik M.Inverse portfolio optimisation under constraints[J].Journal of Asset Management,2008,9(3):239-253.
[14]齐岳,林龙.基于社会责任的多目标证券投资组合模型[J].中国流通经济,2013,27(8):119-123.
[15]Steuer R E,Qi Yue,Hirschberger M.et al.Suitable portfolio investors,nondominated frontier sensitivity,and effects of multiple objectives on portfolio selection[J].Annals of Operations Research,Issue on Financial Optimization,2007,152(1):297-317.