Abstract
Quantifying the risk of unfortunate events occurring, despite limited distributional information, is a basic problem underlying many practical questions. Indeed, quantifying constraint violation probabilities in distributionally robust programming or judging the risk of financial positions can both be seen to involve risk quantification under distributional ambiguity. In this work we discuss worst-case probability and conditional value-at-risk problems, where the distributional information is limited to second-order moment information in conjunction with structural information such as unimodality and monotonicity of the distributions involved. We indicate how exact and tractable convex reformulations can be obtained using standard tools from Choquet and duality theory. We make our theoretical results concrete with a stock portfolio pricing problem and an insurance risk aggregation example.
Similar content being viewed by others
Notes
Note that in the case of structured distributions, one can also derive a multidimensional analog to the Gauss inequality as shown in [23]. In contrast to the present work, which operates on the dual problem (D), the bounds in [23] are produced by operating directly on the primal problem (\(P_{\mathrm {wc}}\)).
The class of \(\gamma \)-monotone distributions defined here can be identified with the class of \((n, \gamma )\)-unimodal distributions discussed in [2, Theorem 3.1.14].
An Intel(R) Core(TM) Xeon(R) CPU E5540 @ 2.53GHz machine.
Note that the integral amounts to \(1-\frac{1}{B(n,\gamma )}\int ^{b_i/q}_0 \lambda ^{n-1}(1-\lambda )^{\gamma -1}~\mathrm {d}\lambda =: 1 - I_{b_i/q}(n,\gamma )\), where \(I_{b_i/q}(n,\gamma )\) is the so-called regularized incomplete beta function, i.e. the cumulative distribution function for the beta distribution with shape parameters \((n,\gamma )\).
References
Bernstein, S.N.: Sur les fonctions absolument monotones. Acta Math. 52(1), 1–66 (1929)
Bertin, E.M.J., Theodorescu, R., Cuculescu, I.: Unimodality of Probability Measures. Mathematics and Its applications. Springer, Berlin (1997)
Bertsimas, D., Popescu, I.: On the relation between option and stock prices: a convex optimization approach. Oper. Res. 50(2), 358–374 (2002)
Delage, E., Ye, Y.: Distributionally robust optimization under moment uncertainty with application to data-driven problems. Oper. Res. 58(3), 595–612 (2010)
Dharmadhikari, S.W., Joag-Dev, K.: Unimodality, Convexity, and Applications, Volume 27 of Probability and Mathematical Statistics. Academic, London (1988)
Embrechts, P., Puccetti, G., Rüschendorf, L.: Model uncertainty and VaR aggregation. J. Bank. Finance 37(8), 2750–2764 (2013)
Gallier, J.: The Schur complement and symmetric positive semidefinite (and definite) matrices. Technical Report, Penn Engineering (2010)
Kiefer, J.: Sequential minimax search for a maximum. Proc. Am. Math. Soc. 4(3), 502–506 (1953)
Lo, A.W.: Semi-parametric upper bounds for option prices and expected payoffs. J. Financ. Econ. 19(2), 373–387 (1987)
Löfberg, J.: Pre- and post-processing sum-of-squares programs in practice. IEEE Trans. Autom. Control 54(5), 1007–1011 (2009)
Nemirovski, A., Shapiro, A.: Convex approximations of chance constrained programs. SIAM J. Optim. 17(4), 969–996 (2006)
Nesterov, Y.: Squared functional systems and optimization problems. In: Frenk, H., Roos, K., Terlaky, T., Zhang, S. (eds.) High Performance Optimization, pp. 405–440. Springer Boston, MA (2000)
Nesterov, Y., Nemirovskii, A.: Interior-Point Polynomial Algorithms in Convex Programming, Volume 13 of Studies in Applied and Numerical Mathematics. SIAM, Philadelphia (1994)
Nocedal, J., Wright, S.J.: Numerical Optimization. Springer Series in Operational Research and Financial Engineering. Springer, New York (2006)
Pestana, D.D., Mendonça, S.: Higher-order monotone functions and probability theory. In: Hadjisavvas, N., Martínez-Legas, J.E., Penot, J-P. (eds.) Generalized Convexity and Generalized Monotonicity, pp. 317–331. Springer, Berlin, Heidelberg (2001)
Phelps, R.R.: Lectures on Choquet’s Theorem, Volume 1757 of Lecture Notes in Mathematics. Springer, Berlin (2001)
Popescu, I.: A semidefinite programming approach to optimal-moment bounds for convex classes of distributions. Math. Oper. Res. 30(3), 632–657 (2005)
Rockafellar, R.T., Uryasev, S.: Optimization of conditional value-at-risk. J. Risk 2, 21–42 (2000)
Rüschendorf, L.: Fréchet-Bounds and Their Applications. Springer, Berlin (1991)
Savage, I.R.: Probability inequalities of the Tchebycheff type. J. Res. Natl. Bur. Stand. B Math. Math. Phys. 65B(3), 211–226 (1961)
Shapiro, A.: On duality theory of conic linear problems. Nonconv. Optim. Appl. 57, 135–155 (2001)
Shapiro, A., Kleywegt, A.: Minimax analysis of stochastic problems. Optim. Methods Softw. 17(3), 523–542 (2002)
Van Parys, B.P.G., Goulart, P.J., Kuhn, D.: Generalized Gauss inequalities via semidefinite programming. Math. Program. 156, 1–32 (2015)
Van Parys, B.P.G., Kuhn, D., Goulart, P.J., Morari, M.: Distributionally robust control of constrained stochastic systems. Technical Report, ETH Zürich (2013)
Van Parys, B.P.G., Ng, B.F., Goulart, P.J., Palacios, R.: Optimal Control for Load Alleviation in Wind Turbines. American Institute of Aeronautics and Astronautics, New York (2014)
Vandenberghe, L., Boyd, S., Comanor, K.: Generalized Chebyshev bounds via semidefinite programming. SIAM Rev. 49(1), 52–64 (2007)
Yamada, Y., Primbs, J.A.: Value-at-risk estimation for dynamic hedging. Int. J. Theor. Appl. Finance 5(4), 333–354 (2002)
Zymler, S., Kuhn, D., Rustem, B.: Distributionally robust joint chance constraints with second-order moment information. Math. Program. 137(1–2), 167–198 (2013)
Zymler, S., Kuhn, D., Rustem, B.: Worst-case value-at-risk of non-linear portfolios. Manag. Sci. 59(1), 172–188 (2013)
Author information
Authors and Affiliations
Corresponding author
Appendices
Equality constrained quadratic programs (QPs)
We will state here a relevant result concerning equality constrained QPs used throughout the rest of this paper. Assume we define a function \(I:\mathbb {R}^d\rightarrow \mathbb {R}\) as follows
with \(A \in \mathbb {R}^{d \times n}\) having full row rank and G positive semidefinite. It is assumed that the function \(x^\top G x + 2 x^\top c\) in bounded from below such that \(I(b)>\infty \). We can now represent the quadratic function I using a dual representation as indicated in the following theorem.
Theorem 5
(Parametric representation of I) The function I is lower bounded by
for all \(T_1 \in \mathbb {S}^{d}\), \(T_2 \in \mathbb {R}^d\) and \(T_3 \in \mathbb {R}\) such that there exist \(\varLambda _1 \in \mathbb {R}^{d\times d}\), \(\varLambda _2 \in \mathbb {R}^d\) with
Moreover, inequality (9) is tight uniformly in \(b\in \mathbb {R}^d\) for some \(T_1\), \(T_2\) and \(T_3\) satisfying condition (10).
Proof
The Lagrangian of the optimization problem defining I(b) is given as
As \( x^\top G x + 2 x^\top c\) is bounded from below on \(\mathbb {R}^n\), we have that for all \(b \in \mathbb {R}^d\) there exists a minimizer \(x^\star \) such that \(I(b) =(x^\star )^\top G x^\star + 2 (x^\star )^\top c + y\) and \(Ax^\star =b\). From the first order optimality conditions for convex QPs [14, Lemma 16.1], we have that \( \min _x \max _\lambda ~ \mathcal {L}(x, \lambda ) = \mathcal L(x^\star , \lambda ^\star ) = \max _\lambda \min _x ~ \mathcal {L}(x, \lambda ) \) where the saddle point \((x^\star , \lambda ^\star )\) is any solution of the linear system
The quadratic optimization problem \(\max _x \, \mathcal L(x, \lambda ^\star )\) admits a maximizer if and only if \((c + A^\top \lambda ^\star )\) is in the range of G. It must thus hold that
Hence when dualizing the problem defining I(b), we get its dual representation \( I(b) = \max _\lambda ~ - \left( c + A^\top \lambda \right) ^\top G^{\dagger } \left( c + A^\top \lambda \right) - 2 \lambda ^\top b + y. \) From equation (11) it follows that \(\lambda ^\star \) is any solution of the linear equation \(b + A G^{\dagger } A^\top \lambda ^\star + A G^{\dagger } c =0\). Therefore there exists an affine \(\lambda ^\star (b) = - \varLambda _1^\star b - \varLambda ^\star _2\) with \(\varLambda _1^\star \in \mathbb {R}^{d\times d}\) and \(\varLambda _2^\star \in \mathbb {R}^d\) such that
From equation (12) it follows that for all b in \(\mathbb {R}^d\) it holds that \( \left( \mathbb {I}_{d} - G G^{\dagger }\right) \left( c-A^\top \varLambda _1^\star b - A^\top \varLambda _2^\star \right) \text{= } \text{0 }. \) We must hence also have that
The dual reprenstation of I(b) guarantees that for all \(\lambda (b) = - \varLambda _1 b - \varLambda _2\) with \(\varLambda _1 \in \mathbb {R}^{d\times d}\) and \(\varLambda _2 \in \mathbb {R}^d\)
Lower bounding the right hand side of the previous inequality with \(b^\top T_1 b + 2 T_2^\top b + T_3 \) yields \(I(b) \ge b^\top T_1 b + 2 T_2^\top b + T_3\) if for all b in \(\mathbb {R}^d\) it holds that
and
After a Schur complement [7, Thm 4.3], we obtain the first part of the theorem
As I(b) is a quadratic function there exist \(T_1^\star \), \(T_2^\star \) and \(T_3^\star \) such that \(I(b)=b^\top T_1^\star b + 2{T_2^\star }^\top b + T_3^\star \). The equations (13) and (14) guarantee [7, Thm 4.3] that
completing the proof. \(\square \)
Proofs
Proposition 2:
Proof
The statement can be proved almost immediately from the definition of convexity. For all \(\theta \in [0, 1]\)
showing convexity of \(L_s\). \(\square \)
Corollary 1:
From Theorem 2, we have that the generating distribution T for \(\alpha \)-unimodal ambiguity sets satisfies
The moment transformations from Theorem 1 become
From Proposition 1, the transformed loss function \(L_s\) required in Theorem 1 can be found as
where
In order to apply Theorem 4, we now need only reformulate the semi-infinite constraint (\(\mathcal C_2\)), i.e. the constraint
Because \(0\in \varXi \) and hence \(b_i > 0\), we have equivalently, for each \(i\in \mathcal I\), and for all \(q \in \mathbb {R}^+\)
which can be seen to reduce to
Defining a new scalar variable \({\tilde{q}}\) and applying the variable substitution \({\tilde{q}}^{\,w} = q\), this can be rewritten as
after multiplying both sides with \({\tilde{q}}^v> 0\). The final result is obtained after the substitution \(b_i^{1/w} \bar{q} = \tilde{q}\).
Corollary 2: The method of proof follows that of Corollary 1, except that we now apply Proposition 3 to generate the transformed loss function \(L_s\).
In this case the loss function L is equivalent to \(L = \ell \circ {\kappa }_{\varXi }\) with \(\ell (t) = \max \{0,t-1\}\). Recalling from Theorem 2 the generating distribution T for \(\alpha \)-unimodal distributions, we set
which is zero for any \(t \le 1\). For \(t\ge 1\), we can evaluate the integral to get
and then set \(L_s(x) = \max _{i\in \mathcal I} f_i(a_i^\top x)\) where each \(f_i(q) := \ell _s(q/b_i)\).
We can now apply Theorem 4 by reformulating the constraint (\(\mathcal C_2\)) for this choice of \(f_i\) for each \(i\in \mathcal I\), resulting in the constraint
because \(0\in \varXi \) and hence \(b_i > 0\). We define a new scalar variable \({\tilde{q}}\) and apply the variable substitution \({\tilde{q}}^{\,w} = q\), resulting in the constraint
after multiplying both sides by \({\tilde{q}}^v> 0\). The final result is obtained after the substitution \(b_i^{1/w} \bar{q} = \tilde{q}\).
Corollary 3:
We follow the same approach as the proof of Corollary 1, but this time use the generating distribution T for \(\gamma \)-monotone distributions from Theorem 3, i.e.
In this case the moment transformations from Theorem 1 become
From Proposition 1, the transformed loss function \(L_s\) required in Theorem 1 become
where
For \(q\ge b_i\), we can use a binomial expansion to evaluate this integral,Footnote 4 obtaining
In order to apply Theorem 4, we now need only reformulate the semi-infinite constraint (\(\mathcal C_2\)). We obtain, for each \(i\in \mathcal I\), the constraint
recalling that \(0\in \varXi \) and hence \(b_i > 0\). We multiply both sides by \(q^{n+\gamma -1}>0\) to produce, for each \(i\in \mathcal I\) the constraint
The final result is obtained after the substitution \(b_i \bar{q} = q\).
Corollary 4:
The method of proof parallels that of Corollary 2, but this time using the generating distribution T for \(\gamma \)-monotone distributions from Theorem 3. In this case we set
which is zero for any \(t \le 1\). For any \(t\ge 1\), using a binomial expansion we can evaluate the integral to get
and then set \(L_s(x) = \max _{i\in \mathcal I} f_i(a_i^\top x)\) where each \(f_i(q) := \ell _s(q/b_i)\). In order to apply Theorem 4, we now need only reformulate the semi-infinite constraint (\(\mathcal C_2\)). We obtain, for each \(i\in \mathcal I\), the constraint
because \(0\in \varXi \) and hence \(b_i > 0\). We multiply both sides by \(q^{n+\gamma -1}>0\) to produce the constraint
The final result is obtained after the substitution \(b_i \bar{q} = \tilde{q}\).
Factor models in insurance
As mentioned in Sect. 1.1, any worst-case CVaR problem can be reduced to a related worst-case expectation problem. We are therefore interested in loss functions of the form \(L(S_d) = \min \left( \max \left( S_d, 0 \right) ,k\right) -\beta \) for \(0\le \beta \le k\). We have that the loss function \(L(S_d)\) can be written as the gauge function \(L(S_d) = \ell \circ \kappa _\varXi (S_d)\) for \(\varXi =\{x\in \mathbb {R}^d|\sum _{i=1}^d x_i \ge 1\}\) and
Recalling from Theorem 2 the generating distribution T for \(\alpha \)-unimodal distributions, we set \( \ell _s(t) = \int _0^{\infty }\ell (\lambda t)T(\mathrm {d}\lambda ) \) which is zero for any \(t \le \beta \). For \(\beta \le t < k\), we can evaluate the integral to get
Similarly for \(t\ge k\), we get
and then set \(L_s(x) = \ell _s(\sum _{i=1}^d x_i)\). In order to apply Theorem 4, we now need only reformulate the semi-infinite constraint (\(\mathcal C_2\)). This can be done using methods analogous to the method described in the proof of Corollary 2, but is omitted here for the sake of brevity. We get finally
Rights and permissions
About this article
Cite this article
Van Parys, B.P.G., Goulart, P.J. & Morari, M. Distributionally robust expectation inequalities for structured distributions. Math. Program. 173, 251–280 (2019). https://doi.org/10.1007/s10107-017-1220-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10107-017-1220-x