Abstract
This paper solves for the robust time-consistent mean–variance portfolio selection problem on multiple risky assets under a principle component stochastic volatility model. The model uncertainty is introduced to the drifts of the risky assets prices and the stochastic eigenvalues of the covariance matrix of asset returns. Using an extended dynamic programming approach, we manage to derive a semi-closed form solution of the desired portfolio via the solution to a coupled matrix Riccati equation. We provide the conditions, under which we prove the existence and the boundedness of the solution to the coupled matrix Riccati equation and derive the value function of the control problem. Moreover, we conduct numerical and empirical studies to perform sensitivity analyses and examine the losses due to ignoring model uncertainty or volatility information.
References
Anderson, E.W., Hansen, L.P., Sargent, T.J.: A quartet of semigroups for model specification, robustness, prices of risk, and model detection. J. Eur. Econ. Assoc. 1, 68–123 (2003)
Andersen, L.B.G., Piterbarg, V.V.: Moment explosions in stochastic volatility models. Financ. Stoch. 11, 29–50 (2007)
Basak, S., Chabakauri, G.: Dynamic mean–variance asset allocation. Rev. Financ. Stud. 23, 2970–3016 (2010)
Biagini, S., Pınar, M.Ç.: The robust Merton problem of an ambiguity averse investor. Math. Financ. Econ. 11, 1–24 (2017)
Bergen, V., Escobar, M., Rubtsov, A., Zagst, R.: Robust multivariate portfolio choice with stochastic covariance in the presence of ambiguity. Quant. Financ. 18, 1265–1294 (2018)
Björk, T., Murgoci, A., Zhou, X.Y.: Mean–variance portfolio optimization with state-dependent risk aversion. Math. Financ. 24, 1–24 (2014)
Björk, T., Khapko, M., Murgoci, A.: On time-inconsistent stochastic control in continuous time. Financ. Stoch. 21, 331–360 (2017)
Chiu, M.C., Wong, H.Y.: Mean-variance portfolio selection with correlation risk. J. Comput. Appl. Math. 263, 432–444 (2014)
Chiu, M.C., Wong, H.Y.: Optimal investment for insurers with correlation risk: risk aversion and investment horizon. IMA J. Manag. Math. 29, 207–227 (2018)
Cox, J.C., Ingersoll, J.E., Ross, S.A.: A theory of the term structure of interest rates. Econometrica 53, 385–407 (1985)
Ellsberg, D.: Risk, ambiguity, and the savage axioms. Q. J. Econ. 75, 643–669 (1961)
Escobar, M., Götz, B., Seco, L., Zagst, R.: Pricing a CDO on stochastically correlated underlyings. Quant. Financ. 10, 265–277 (2010)
Fouque, J.-P., Pun, C.S., Wong, H.Y.: Portfolio optimization with ambiguous correlation and stochastic volatilities. SIAM J. Control Optim. 54, 2309–2338 (2016)
Hambly, B.M., Kolliopoulos, N.: Stochastic evolution equations for large portfolios of stochastic volatility models. SIAM J. Financ. Math. 8, 962–1014 (2017)
Han, B., Pun, C.S., Wong, H.Y.: Robust time-inconsistent stochastic linear-quadratic control: an open-loop approach. https://ssrn.com/abstract=3167662 (2018)
Han, B., Pun, C.S., Wong, H.Y.: Robust mean–variance portfolio selection with state-dependent ambiguity and risk aversion: a closed-loop approach. https://ssrn.com/abstract=3306305 (2018)
Heston, S.L.: A closed-form solution for options with stochastic volatility with applications to bond and currency options. Rev. Financ. Stud. 6, 327–343 (1993)
Hu, Y., Jin, H., Zhou, X.Y.: Time-inconsistent stochastic linear-quadratic control. SIAM J. Control Optim. 50, 1548–1572 (2012)
Hu, Y., Jin, H., Zhou, X.Y.: Time-inconsistent stochastic linear-quadratic control: characterization and uniqueness of equilibrium. SIAM J. Control Optim. 55, 1261–1279 (2017)
Knight, F.H.: Risk, Uncertainty, and Profit. Houghton Mifflin, New York (1921)
Li, Z.F., Zeng, Y., Lai, Y.Z.: Optimal time-consistent investment and reinsurance strategies for insurers under Heston’s SV model. Insur. Math. Econ. 51, 191–203 (2012)
Maenhout, P.J.: Robust portfolio rules and asset pricing. Rev. Financ. Stud. 17, 951–983 (2004)
Markowitz, H.: Portfolio selection. J. Financ. 7, 77–91 (1952)
Merton, R.C.: Optimum consumption and portfolio rules in a continuous-time model. J. Econ. Theory 3, 373–413 (1971)
Papavassilopoulos, G.P., Cruz, J.B.: On the existence of solutions to coupled matrix Riccati differential equations in linear quadratic Nash games. IEEE T. Autom. Control 24, 127–129 (1979)
Pedersen, J.L., Peskir, G.: Optimal mean–variance portfolio selection. Math. Financ. Econ. 11, 137–160 (2017)
Pedersen, J.L., Peskir, G.: Optimal mean–variance selling strategies. Math. Financ. Econ. 10, 203–220 (2016)
Pun, C.S.: Robust time-inconsistent stochastic control problems. Automatica 94, 249–257 (2018)
Pun, C.S.: \(G\)-expected utility maximization with ambiguous equicorrelation. SSRN: https://papers.ssrn.com/abstract=3276001 (2018)
Pun, C.S., Wong, H.Y.: Robust investment–reinsurance optimization with multiscale stochastic volatility. Insur. Math. Econ. 62, 245–256 (2015)
Rieder, U., Wopperer, C.: Robust consumption–investment problems with random market coefficients. Math. Financ. Econ. 6, 295–311 (2012)
Sun, Z., Zheng, X., Zhang, X.: Robust optimal investment and reinsurance of an insurer under variance premium principle and default risk. J. Math. Anal. Appl. 446, 1666–1686 (2017)
Wald, A.: Statistical decision functions which minimize the maximum risk. Ann. Math. 46, 265–280 (1945)
Wei, J., Wang, T.: Time-consistent mean–variance asset-liability management with random coefficients. Insur. Math. Econ. 77, 84–96 (2017)
Yan, T., Wong, H.Y.: Open-loop equilibrium strategy for mean–variance portfolio problem under stochastic volatility. Automatica 107, 211–223 (2019)
Yi, B., Li, Z., Viens, F., Zeng, Y.: Robust optimal control for an insurer with reinsurance and investment under Heston’s stochastic volatility model. Insur. Math. Econ. 53, 601–614 (2013)
Yong, J., Zhou, X.Y.: Stochastic controls: Hamiltonian systems and HJB equations. Springer, New York (1999)
Zhou, X.Y., Li, D.: Continuous-time mean–variance portfolio selection: a stochastic LQ framework. Appl. Math. Optim. 42, 19–33 (2000)
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 was supported by Ministry of Education, Singapore under AcRF Tier 2 Grant (Number: MOE2017-T2-1-044).
Appendices
Appendix A. Proof of Lemma 1
Proof
We first prove the second inequality. By Lemma 3.2 in [14], we have \({\mathbb {E}}^{{\mathbb {P}}}\left[ \sup _{0\le t \le T}\Vert {\varvec{\varLambda }}_t\Vert ^{4p} \right] < \infty \) for any \(p\ge 1\). Then, by the Hölder’s inequality, we have
where the last equality holds since we can define a new measure using the Radon-Nikodym derivative \(\mathcal {E}_T\left( 4{\varvec{q}}^{\varvec{S}}\cdot {\varvec{W}}^{\varvec{S}}\right) \mathcal {E}_T\left( 4{\varvec{q}}^{{\varvec{\varLambda }}} \cdot {\varvec{W}}^{\varvec{\varLambda }}\right) \) based on the Novikov’s condition (7).
Moreover, due to the condition (1) in Definition 1, we have \({\mathbb {E}}^{\mathbb {P}}\left[ \left( \int _0^T \Vert {\varvec{q}}_s\Vert ^2ds \right) ^p\right] < \infty \) for any \(p\ge 1\). By mimicking the steps in (26), we have \({\mathbb {E}}^{\mathbb {Q}}\left[ \left( \int _0^T \Vert {\varvec{q}}_s\Vert ^2ds \right) ^p\right] < \infty \) for any \(p\ge 1\). Therefore, by (6), (8), (9) and Doob’s martingale maximal inequality, there exist constants \(C_1, C_2, C_3 > 0\) such that
\(\square \)
Appendix B. Proof of Lemma 4
Proof
We only prove that \(\mathcal {A}^{{\varvec{u}}, {\varvec{q}}} g^2(t,X_t,{\varvec{\varLambda }}_t) \in {\mathbb {L}}^1_{\mathcal {F}}(0,\, T;\, {\mathbb {R}}, \, {\mathbb {Q}})\), \({{\varvec{\sigma }}^{X}({\varvec{u}}_t, {\varvec{\varLambda }}_t)}' \frac{\partial g^2}{\partial x}(t, X_t, {\varvec{\varLambda }}_t) \in {\mathbb {L}}^2_{\mathcal {F}}(0,\, T; \, {\mathbb {R}}^n,\, {\mathbb {Q}})\); others can be proved with similar arguments. For simplicity, we omit the argument t of functions g and \(g_1\) in our proof.
When conditions in Lemma 3 hold, by Lemmas 1 and 3, there exist three constants \(C_1, C_2, C_3>0\) such that
Again, by Lemma 2.2, it is easy to verify that \(\mathcal {A}^{{\varvec{u}}, {\varvec{q}}} g^2(t,X_t,{\varvec{\varLambda }}_t) \in {\mathbb {L}}^1_{\mathcal {F}}(0,\, T;\, {\mathbb {R}}, \, {\mathbb {Q}})\).
Similarly, there exist two constants \(C_4, C_5>0\) such that
whose upper bound is finite by noting the construction of the admissible control-measure policy and Lemma 1. \(\square \)
Appendix C. Proof of Proposition 1
Proof
We ought to prove that under the assumptions in this proposition, \(({\varvec{u}}^*, {\varvec{q}}^*)\) given by (22) is an admissible control-measure policy in Definition 1. The rest is a direct consequence of Theorem 1 and Lemmas 2, 3, and 4.
We note that
which is the product of the Laplace transform of the integrated CIR process. We first discuss the boundedness of \(V_{2i}(t)\) on [0, T]. By the Proposition in [25], we obtain the following results under the conditions in Lemma 3:
-
(1)
Under condition (i), \(|V_{2i}(t)| \le \eta ^i_1 +\frac{1}{\beta ^i(\frac{2}{\alpha ^i}-T)}\);
-
(2)
Under condition (ii), \(|V_{2i}(t)| \le \eta ^i_2 +\frac{\sqrt{\varDelta ^i}}{\beta ^i(1-\frac{\eta ^i_1}{\eta ^i_2}e^{\sqrt{\varDelta ^i}T})}\);
-
(3)
Under condition (iii), \(|V_{2i}(t)| \le -\frac{\alpha ^i}{2\beta ^i} + \frac{\sqrt{|\varDelta ^i|}}{2\beta ^i} \tan \left( \frac{\sqrt{|\varDelta ^i|}}{2}T + \arctan \frac{\alpha ^i}{\sqrt{|\varDelta ^i|}}\right) \),
for any \(t\in [0,T]\) and \(i = 1, 2, \ldots , n\). We use function \(U_i(T)\) to denote the bound of \(V_{2i}(t)\) on [0, T]. Define \(\bar{U}_i(T)=\frac{8\xi ^2\delta _i^2}{(\xi +\gamma )^2} + 8\xi ^2\sigma _i^2(1-\rho _i^2)U_i^2(T)\). By [2], \({\mathbb {E}}^{{\mathbb {P}}}\left[ \exp \left( 8\int _0^T \Vert {\varvec{q}}_s^*\Vert ^2 ds\right) \right] <\infty \) if \(k_i^2 - 2 \bar{U}_i(T)\sigma _i^2 \ge 0\) for \(i=1, 2, \ldots , n\).
The assumptions on T in this proposition are imposed such that both conditions in Lemma 3 and \(k_i^2 - 2 \bar{U}_i(T)\sigma _i^2 \ge 0\) are satisfied and thus the first condition of Definition 1 is satisfied. We remark here that one can easily verify that all the upper bounds of T in the assumptions are positive.
Since \({\varvec{u}}^*\) is a deterministic continuous function on \(t \in [0,T]\), by Lemma 1, \(({\varvec{u}}^*, {\varvec{q}}^*)\) is an admissible control-measure policy. \(\square \)
Rights and permissions
About this article
Cite this article
Yan, T., Han, B., Pun, C.S. et al. Robust time-consistent mean–variance portfolio selection problem with multivariate stochastic volatility. Math Finan Econ 14, 699–724 (2020). https://doi.org/10.1007/s11579-020-00271-0
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11579-020-00271-0
Keywords
- Time-inconsistency
- Dominated model uncertainty
- Mean–variance portfolio selection
- Stochastic covariance matrix
- Principal component stochastic volatility model
- Hamilton–Jacobi–Bellman–Isaacs equations