Abstract
We consider the factor-risk-constrained mean-variance portfolio-selection (MVPS) problem that allows managers to construct portfolios with desired factor-risk characteristics. Its optimization model is a non-convex quadratically constrained quadratic program that is known to be NP-hard. In this paper, we investigate the new global algorithm for factor-risk-constrained MVPS problem based on the successive convex optimization (SCO) method and the semi-definite relaxation (SDR) with a second-order cone (SOC) constraint. We first develop an SCO algorithm and show that it converges to a KKT point of the problem. We then develop a new global algorithm for factor-risk-constrained MVPS, which integrates the SCO method, the SDR with an SOC constraint, the branch-and-bound framework and the adaptive branch-and-cut rule for factor-related variables, to find a globally optimal solution to the underlying problem within a pre-specified \(\epsilon \)-tolerance. We establish the global convergence of the proposed algorithm and its complexity. Preliminary numerical results demonstrate the effectiveness of the proposed algorithm in finding a globally optimal solution to medium- and large-scale instances of factor-risk-constrained MVPS.
Similar content being viewed by others
Notes
All the data used in Sect. 5 can be downloaded on https://github.com/hezhiluo/MVPS.
In our numerical experiments, both the SOCR subproblem in BB-SOC and the SDR subproblem in BB-SDRACS are solved by SeDuMi [52].
Source code for IPOPT can be downloaded from: https://github.com/coin-or/Ipopt.
References
Audet, C., Hansen, P., Jaumard, B., Savard, G.: A branch and cut algorithm for nonconvex quadratically constrained quadratic programming. Math. Program. 87, 131–152 (2000)
Avriel, M.: Nonlinear Programming: Analysis and Methods. Prentice-Hall Co., Inc., New Jersey (1976)
Bertsimas, D., Lauprete, G.J., Samarov, A.: Shortfall as a risk measure: Properties, optimization and applications. J. Econ. Dyn. Control 28, 1353–1381 (2004)
Bryant, F.B., Yarnold, P.R.: Principal Components Analysis and Exploratory and Confirmatory Factor Analysis. In: Reading and Understanding Multivariate Statistics. American Psychological Association, Washington, DC (1995)
Burer, S., Vandenbussche, D.: A finite branch-and-bound algorithm for nonconvex quadratic programming via semidefinite relaxations. Math. Program. 113(2), 259–282 (2008)
Burer, S., Vandenbussche, D.: Globally solving box-constrained nonconvex quadratic programs with semidefinite-based finite branch-and-bound. Comput. Optim. Appl. 43(2), 181–195 (2009)
Burmeister, E., Ross, S.A.: Using macroeconomic factors to control portfolio risk. Technical Report, Duke University (2003)
Chen, J., Burer, S.: Globally solving nonconvex quadratic programming problems via completely positive programming. Math. Program. Comput. 4, 33–52 (2012)
Connor, G.: The three types of factor models: a comparison of their explanatory power. Financ. Anal. J. 51(3), 42–46 (1995)
Ding, X.D., Luo, H.Z., Wu, H.X., Liu, J.Z.: An efficient global algorithm for worst-case linear optimization under uncertainties based on nonlinear semidefinite relaxation. Comput. Optim. Appl. 80(1), 89–120 (2021)
Fama, E.F., French, K.R.: Common risk factors in the returns on stock and bonds. J. Financial Econom. 33(1), 3–56 (1993)
Fama, E.F., French, K.R.: Size and book-to-market factors in earnings and returns. J. Finance 50(1), 131–155 (1995)
Fama, E.F., French, K.R.: A five-factor asset pricing model. J. Financial Econom. 116(1), 1–22 (2015)
Floudas, C.A., Visweswaran, V.: Quadratic optimization. Handbook of Global Optimization, R.Horst and P.M. Pardalos, Eds., Kluwer Academic Publishers, pages 217–270 (1994)
Gould, N.I.M., Toint, P.L.: Numerical methods for large-scale non-convex quadratic programming In: Trends in Industrial and Applied Mathematics (Amritsar, Springer, Boston, 2001). Appl. Optim. 72, 149–179 (2002)
Grinold, R., Kahn, R.: Active Portfolio Management: A Quantitative Approach for Producing Superior Return and Controlling Risk. McGraw-Hill, New York (1999)
Jeff Hong, L., Yang, Yi., Zhang, L.W.: Sequential convex approximations to joint chance constrained programs: A monte carlo approach. Oper. Res. 59(3), 617–630 (2011)
IBM ILOG CPLEX. IBM ILOG CPLEX 12.6 User’s Manual for CPLEX (2013). http://www.cplex.com
Konno, H., Yamazaki, H.: Mean-absolute deviation portfolio optimization model and its applications to tokyo stock market. Manage. Sci. 37, 519–531 (1991)
Kurth, A., Tasche, D.: Credit risk contributions to value-at-risk and expected shortfall. Risk 16, 84–88 (2003)
Lanckriet, G.R., Sriperumbudur, B.K.: On the convergence of the concave-convex procedure. Adv. Neural. Inf. Process. Syst. 22, 1759–1767 (2009)
Le Thi, H.A., Huynh, V.N., Pham Dinh, T.: DC programming and DCA for general DC programs. In: Advanced Computational Methods for Knowledge Engineering, pages 15–35. Springer International Publishing (2014)
Le Thi, H.A., Pham Dinh, T.: Solving a class of linearly constrained indefinite quadratic problems by D.C. algorithms. J. Global Optim. 11, 253–285 (1997)
Li, D., Ng, W.L.: Optimal dynamic portfolio selection: multi-period mean-variance formulation. Math. Finance 10, 387–406 (2004)
Li, Q., Bai, Y.Q.: Optimal trade-off portfolio selection between total risk and maximum relative marginal risk. Optim. Methods Softw. 31(4), 681–700 (2016)
Li, Q., Bai, Y.Q., Yan, X., Zhang, W.: Portfolio selection with the effect of systematic risk diversification: formulation and accelerated gradient algorithm. Optim. Methods Softw. 34(3), 612–633 (2019)
Li, Y.J., Zhu, S.S., Li, D.H., Li, D.: Active allocation of systematic risk and control of risk sensitivity in portfolio optimization. Eur. J. Oper. Res. 228(3), 556–570 (2013)
Linderoth, J.: A simplicial branch-and-bound algorithm for solving quadratically constrained quadratic programs. Math. Program., Ser. B 103, 251–282 (2005)
Lipp, T., Boyd, S.: Variations and extension of the convex-concave procedure. Optim. Eng. 17(2), 263–287 (2016)
Loridan, P.: Necessary conditions for \(\epsilon \)-optimality. Math. Program. Stud. 19, 140–152 (1982)
Lu, C., Liu, Y.F.: An efficient global algorithm for single-group multicast beamforming. IEEE Trans. Signal Process. 65(14), 3761–3774 (2017)
Lu, C., Deng, Z.B., Zhou, J., Guo, X.L.: A sensitive-eigenvector based global algorithm for quadratically constrained quadratic programming. J. Global Optim. 73(2), 371–388 (2019)
Lu, C., Liu, Y.F., Zhou, J.: An enhanced SDR based global algorithm for nonconvex complex quadratic programs with signal processing applications. IEEE Open J. Signal Process. 1, 120–134 (2020)
Luo, H.Z., Ding, X.D., Peng, J.M., Jiang, R.J., Li, D.: Complexity results and effective algorithms for the worst-case linear optimization under uncertainties. INFORMS J. Comput. 33(1), 180–197 (2021)
Luo, H.Z., Bai, X.D., Lim, G., Peng, J.M.: New global algorithms for quadratic programming with a few negative eigenvalues based on alternative direction method and convex relaxation. Math. Program. Comput. 11(1), 119–171 (2019)
Luo, H.Z., Chen, S.K., Wu, H.X.: A new branch-and-cut algorithm for non-convex quadratic programming via alternative direction method and semidefinite relaxation. Numer. Algorithms 88(2), 993–1024 (2021)
Luo, H.Z., Chen, Y.Y., Zhang, X.Y., Li, D., Wu, H.X.: Effective algorithms for optimal portfolio deleveraging problem with cross impact. arXiv:2012.07368v2, pages 1–50 (2021)
Markowitz, H.: Portfolio selection. J. Finance 7(1), 77–91 (1952)
Morgan, J.P.: Credit Metrics Technical Document, Risk Metrics. J.P. Morgan, New York (1997)
Palacios-Gomez, F., Lasdon, L., Enquist, M.: Nonlinear optimization by successive linear programming. Manage. Sci. 28(10), 1106–1120 (1982)
Pham Dinh, T., Le Thi, H.A.: Recent advances in DC programming and DCA. In: Transactions on Computational Intelligence XIII, 1–37. Springer, Berlin, Heidelberg (2014)
Philippe, J.: Value at Risk: The New Benchmark for Controlling Market Risk. Irwin Professional Publishing, Chicago (1996)
Philippe, J.: Value at Risk: The New Benchmark for Managing Financial Risk, 3rd edn. McGraw-Hill, New York (2006)
Rockafellar, R.T., Uryasev, S.: Optimization of conditional value-at-risk. J. Risk 2, 21–41 (2000)
Rockafellar, R.T., Wets, R.J.-B.: Variational Analysis. Springer-Verlag, New York (1998)
Ross, S.A.: The arbitrage theory of capital asset pricing. J. Econ. Theory 13, 341–360 (1976)
Saxena, A., Bonami, P., Lee, J.: Convex relaxation of nonconvex mixed integer quadratically constrained programs: Projected formulations. Math. Program. 130, 359–413 (2011)
Sharpe, W.F.: A simplified model for portfolio analysis. Manage. Sci. 9(2), 277–293 (1963)
Sharpe, W.F.: Capital asset prices: a theory of market equilibrium under conditions of risk. J. Finance 19(3), 425–442 (1964)
Sharpe, W.F.: Budgeting and monitoring pension fund risk. Financ. Anal. J. 58, 74–86 (2002)
Solodov, M.V.: On the sequential quadratically constrained quadratic programming methods. Math. Oper. Res. 29(1), 64–79 (2004)
Sturm, J.F.: Using SeDuMi 1.02, a MATLAB toolbox for optimization over symmetric cones. Optim. Methods Softw. 11, 625–653 (1999)
Wächter, A., Biegler, L.T.: Line search filter methods for nonlinear programming: Local convergence. SIAM J. Optim. 16(1), 32–48 (2005)
Wächter, A., Biegler, L.T.: Line search filter methods for nonlinear programming: Motivation and global convergence. SIAM J. Optim. 16(1), 1–31 (2005)
Wächter, A., Biegler, L.T.: On the Implementation of a primal-dual interior point filter line search algorithm for large-scale nonlinear programming. Math. Program. 106(1), 25–57 (2006)
Ye, Y.: On the complexity of approximating a KKT point of quadratic programming. Math. Program. 80, 195–211 (1998)
Zhu, S.S., Li, D., Wang, S.Y.: Risk control over bankruptcy in dynamic portfolio selection: a generalized mean-variance formulation. IEEE T. Automat. Contr. 49(3), 447–457 (2004)
Zhu, S.S., Cui, X.T., Sun, X.L., Li, D.: Factor-risk-constrained mean-variance portfolio selection: Formulation and global optimization solution approach. J. Risk 14(2), 51–89 (2011)
Zhu, S.S., Li, D., Sun, X.L.: Portfolio selection with marginal risk control. J. Comput. Finance 14(1), 3–28 (2010)
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This work is jointly supported by the National Natural Science Foundation of China (NSFC) [Grants 11871433 and 11371324] and the Zhejiang Provincial NSFC [grants LZ21A010003, LY18A010011 and LQ17A010009].
Appendix: proof of prosition 2
Appendix: proof of prosition 2
In this appendix, we present the proof of Prosition 2, which needs the following two lemmas.
Lemma 8
Let \(\{x^k \}\) and \(\{\xi ^k \}\) be the infinite sequences generated by Algorithm 2. Then,
for all k, where \(\delta _{\min }=\min \{\delta _j,j\in J\}\).
Proof
We first rewrite problem (27) as the following
Note that two problems (27) and (40) are equivalent in the sense that they have the same optimal value and the same solution set. From Step 1, \((x^{k+1},\nu ^{k+1})\) is the optimal solution of problem (27). Thus, \( x^{k+1}\) is the optimal solution of problem (40), and
Note that \(x^{k}\) is a feasible solution of problem (40) and \(\tau _k>0\). Thus,
By noting that
and that \(\xi ^k=Mx^k\) for all k, we yield \({\tilde{g}}_j(x^k;\xi ^k)=g_j(x^k)\) for all k. It then follows from (42) that
On the other hand, note that \(\xi ^{k+1}=Mx^{k+1}\), we have from (41) that \(v^{k+1}\ge 0\) and
where \(\delta _{\min }=\min \{\delta _j,j\in J\}\). The above inequality yields
which further implies
It then follows from (44) and (45) that for all k, one has
The proof of the lemma is finished. \(\square \)
Lemma 9
Let the sequence \(\{\tau _k\}\) be generated by Algorithm 2. Suppose that Assumption 1 holds at any \(x\in {{{\mathcal {D}}}}\) with \(g(x)\ge 0\). Then, there exists some index \(k_0\) such that \(\tau _k= \tau _{k_0}\) for all \(k\ge k_0\).
Proof
Suppose, by contradiction, that \(\tau _k\rightarrow +\infty \) as \(k\rightarrow \infty \). Then (31) implies that \(\tau _k<\gamma _k=\min \{\Vert \xi ^{k+1}-\xi ^k\Vert ^{-1},\Vert \eta ^{k+1}\Vert _1+\delta \}\) happens an infinite number of times. It then further follows that there exists a subsequence of iteration indices \(\{k_i\}\), such that
Taking a further subsequence, if necessary, we can assume that \(\{x^{k_i+1}\}\rightarrow {\bar{x}}\) as \(i\rightarrow \infty \). By Step 1, since \((x^{k_i+1},\nu ^{k_i+1})\) is the optimal solution of convex problem (27) with \(k=k_i\) and \((\eta ^{k_i+1},\mu ^{k_i+1})\) is the associated Lagrange multipliers, we have from the necessary optimality condition (29) that
Since \(\xi ^{k_i+1}=Mx^{k_i+1}\) by Step 1, we have
If \(g({\bar{x}})<0\), then when \(k_i\) is sufficiently large, one has
Thus, by (47), \(\eta _j^{k_i+1}=0\), \(j\in J\) and then \(\Vert \eta ^{k_i+1}\Vert _1=0\), which contradicts \(\Vert \eta ^{k_i+1}\Vert _1=+\infty \) as \(i\rightarrow \infty \). Therefore, \(g({\bar{x}})\ge 0\). Let
Then,
By the definition of \({{{\mathcal {N}}}}_{{{{\mathcal {D}}}}}(x^{k_i+1})\), the first inclusion of (47) yields
Note that since \(\xi ^{k+1}=Mx^{k+1}\) by Step 1, we obtain
which, by (46) and \(\{x^{k_i+1}\}\rightarrow {\bar{x}}\) as \(i\rightarrow \infty \), implies
By dividing the inequality (49) by \(\Vert \eta ^{k_i+1}\Vert _1\) and letting \(i\rightarrow \infty \), we obtain
which in turn implies that
This, by \(\sum _{j\in J({\bar{x}})}{\bar{\eta }}_j=1\), contradicts Condition (9) holds at \({\bar{x}}\) with \(g({\bar{x}})\ge 0\). \(\square \)
Proof of Prosition 2
By Lemma 8, \(\lim _{k\rightarrow \infty } \Vert \xi ^{k+1}-\xi ^{k}\Vert =0\). Then from (30), \(\gamma _k=\Vert \eta ^{k+1}\Vert _1+\delta \) when k is sufficiently large. By Lemma 9, there exists some index \(k_0\) such that \(\tau _k= \tau _{k_0}\) for all \(k\ge k_0\). Hence, from (31), \(\tau _k\ge \Vert \eta ^{k+1}\Vert _1+\delta \) for all k sufficiently large. In view of the second relation of (29), it follows that \(\mu ^{k+1}> 0\), thus, \(\nu ^{k+1} = 0\) when k is sufficiently large. From the last relation of (29), we have that \({\tilde{g}}_j(x^{k+1};\xi ^{k})\le 0\) for \(j\in J\) when k is sufficiently large. Note that since \(\xi ^{k+1}=Mx^{k+1}\) by Step 1, from (43) we have that \({\tilde{g}}_j(x^{k+1};\xi ^{k})=g_j(x^{k+1})+\delta _j\Vert \xi ^{k+1}-\xi ^{k}\Vert ^2\). Since \(\lim _{k\rightarrow \infty } \Vert \xi ^{k+1}-\xi ^{k}\Vert =0\), we obtain \(\lim _{k\rightarrow \infty }{\tilde{g}}_j(x^{k+1};\xi ^{k})=g_j({\bar{x}})\). Consequently, \(g_j({\bar{x}})\le 0\) and so \({\bar{x}}\in {{\mathcal {F}}}\). \(\square \)
Rights and permissions
Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Wu, H., Luo, H., Zhang, X. et al. A new global algorithm for factor-risk-constrained mean-variance portfolio selection. J Glob Optim 87, 503–532 (2023). https://doi.org/10.1007/s10898-022-01218-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10898-022-01218-z