An efficient optimization approach for a cardinality-constrained index tracking problem

被引:44
作者
Xu, Fengmin [1 ]
Lu, Zhaosong [2 ]
Xu, Zongben [1 ]
机构
[1] Xi An Jiao Tong Univ, Sch Math & Stat, Xian 710049, Peoples R China
[2] Simon Fraser Univ, Dept Math, Burnaby, BC V5A 1S6, Canada
基金
加拿大自然科学与工程研究理事会;
关键词
index tracking; cardinality constraint; nonmonotone projected gradient method; PORTFOLIO OPTIMIZATION; ERROR; MINIMIZATION; SPARSE;
D O I
10.1080/10556788.2015.1062891
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
In the practical business environment, portfolio managers often face business-driven requirements that limit the number of constituents in their tracking portfolio. A natural index tracking model is thus to minimize a tracking error measure while enforcing an upper bound on the number of assets in the portfolio. In this paper we consider such a cardinality-constrained index tracking model. In particular, we propose an efficient nonmonotone projected gradient (NPG) method for solving this problem. At each iteration, this method usually solves several projected gradient subproblems. We show that each subproblem has a closed-form solution, which can be computed in linear time. Under some suitable assumptions, we establish that any accumulation point of the sequence generated by the NPG method is a local minimizer of the cardinality-constrained index tracking problem. We also conduct empirical tests to compare our method with the hybrid evolutionary algorithm [P.R. Torrubiano and S. Alberto. A hybrid optimization approach to index tracking. Ann Oper Res. 166(1) (2009), pp. 57-71] and the hybrid half thresholding algorithm [F. Xu, Z. Xu and H Xue. Sparse index tracking: an L-1/2 regularization based model and solution, Submitted, 2012] for index tracking. The computational results demonstrate that our approach generally produces sparse portfolios with smaller out-of-sample tracking error and higher consistency between in-sample and out-of-sample tracking errors. Moreover, our method outperforms the other two approaches in terms of speed.
引用
收藏
页码:258 / 271
页数:14
相关论文
共 29 条
[1]   Optimal hedging using cointegration [J].
Alexander, C .
PHILOSOPHICAL TRANSACTIONS OF THE ROYAL SOCIETY A-MATHEMATICAL PHYSICAL AND ENGINEERING SCIENCES, 1999, 357 (1758) :2039-2058
[2]   Tracking error and tactical asset allocation [J].
Ammann, M ;
Zimmermann, H .
FINANCIAL ANALYSTS JOURNAL, 2001, 57 (02) :32-43
[3]  
[Anonymous], QUANT FINANC
[4]   Nonmonotone projected gradient methods based on barrier and Euclidean distances [J].
Auslender, Alfred ;
Silva, Paulo J. S. ;
Teboulle, Marc .
COMPUTATIONAL OPTIMIZATION AND APPLICATIONS, 2007, 38 (03) :305-327
[5]   2-POINT STEP SIZE GRADIENT METHODS [J].
BARZILAI, J ;
BORWEIN, JM .
IMA JOURNAL OF NUMERICAL ANALYSIS, 1988, 8 (01) :141-148
[6]   An evolutionary heuristic for the index tracking problem [J].
Beasley, JE ;
Meade, N ;
Chang, TJ .
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2003, 148 (03) :621-643
[7]   Nonmonotone spectral projected gradient methods on convex sets [J].
Birgin, EG ;
Martínez, JM ;
Raydan, M .
SIAM JOURNAL ON OPTIMIZATION, 2000, 10 (04) :1196-1211
[8]   Sparse and stable Markowitz portfolios [J].
Brodie, Joshua ;
Daubechies, Ingrid ;
De Mol, Christine ;
Giannone, Domenico ;
Loris, Ignace .
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA, 2009, 106 (30) :12267-12272
[9]  
Buckney I.R.C., 1998, INT J APPL THEORETIC, V1, P315
[10]   Mixed-integer programming approaches for index tracking and enhanced indexation [J].
Canakgoz, N. A. ;
Beasley, J. E. .
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2009, 196 (01) :384-399