Abstract
This paper presents a novel framework for optimizing portfolios using distribution dependent thresholds in Omega ratio to control the downside risk. Portfolios resulting from the maximization of the classical Omega ratio simultaneously maximize the probability of superior performance compared to a threshold point set by an investor and minimize the probability of a worse performance compared to the same threshold. However, there is no mandatory rule or mechanism to choose this threshold point in the Omega ratio optimization model yet. In this paper, we redefine the Omega ratio for a loss averse investor by taking the distribution dependent threshold point as the conditional value-at-risk at an \(\alpha \) confidence level (\( {\mathrm{CVaR}_{\alpha }}\)) of the benchmark market. The \(\alpha \)-value reflects the attitude of an investor towards losses. We then embed this new Omega-\( {\mathrm{CVaR}_{\alpha }}\) model in a robust portfolio optimization framework and present its worst case analysis under three uncertainty sets. The robustness is introduced both in the Omega measure and the \( {\mathrm{CVaR}_{\alpha }}\) measure. We show that the worst case Omega-\( {\mathrm{CVaR}_{\alpha }}\) robust optimization models are linear programs for mixed and box uncertainty sets and a second order cone program under ellipsoidal sets, and hence tractable in all three cases. We conduct a comprehensive empirical investigation of the classical \( {\mathrm{CVaR}_{\alpha }}\) model, the STARR\(_{\alpha }\) model, the Omega-\( {\mathrm{CVaR}_{\alpha }}\) model, and robust Omega-\( {\mathrm{CVaR}_{\alpha }}\) model under a mixed uncertainty set for listed stocks of the S&P 500. The optimal portfolios resulting from the Omega-\( {\mathrm{CVaR}_{\alpha }}\) model exhibit a superior performance compared to the classical \( {\mathrm{CVaR}_{\alpha }}\) model in the sense of higher expected returns, Sharpe ratios, modified Sharpe ratios, and lesser losses in terms of \({\mathrm{VaR}_{\alpha }}\) and \( {\mathrm{CVaR}_{\alpha }}\) values. The robust Omega-\( {\mathrm{CVaR}_{\alpha }}\) model under mixed uncertainty set is shown to dominate the Omega-\( {\mathrm{CVaR}_{\alpha }}\) model in terms of all performance measures. Furthermore, both the Omega-\( {\mathrm{CVaR}_{\alpha }}\) and robust Omega-\( {\mathrm{CVaR}_{\alpha }}\) model under a mixed uncertainty set yield significantly lower risk compared to STARR\(_{\alpha }\) model in terms of \(\mathrm{CVaR}_{\alpha }\) and variance values.
Similar content being viewed by others
Notes
To be precise, variance is the second central moment of the distribution. However, it can be computed by the first two moments.
Note that the underlying distribution is the loss distribution in which expected losses below the threshold indicate smaller losses than the \({\mathrm{CVaR}_{\alpha }}\).
With respect to the definition of the Omega ratio, the value-at-risk is not the natural choice for the threshold. In fact, the Omega ratio for the value of risk at a certain confidence level is constant (when distribution to compute Omega ratio is similar to the distribution to compute value-at-risk) due to the definition of the value-at-risk.
Thus, in this framework, we are able to reduce the sensitivity of the Omega ratio to its threshold point, which now is also chosen by the robust optimization approach.
For computational purposes, we analyze the worst case of Omega-\( {\mathrm{CVaR}_{\alpha }}\) under the mixed uncertainty set.
Heavy tails and dispersion around the mean return (as, at mean return, the Omega ratio is a constant) are controlled by setting upper bounds on risk measures \( {\mathrm{CVaR}_{\alpha }}\), minimax (Young 1998), and semi-mean absolute deviation SemiMAD (Ogryczak and Ruszczynski 1999) in constraints in the Omega ratio model while maintaining linearity in the resulting three hybrid models.
A scenario is a particular realization of the uncertain data.
In other words, it conveys that value of \( {\mathrm{CVaR}_\alpha }\) (which accounts for the right tail of loss distribution of z) is larger than the minimum expected value of the loss function (which accounts for the left tail of loss distribution of x)
We take z as all those stocks listed on the S&P 500 whose monthly return data is available for more than 10 years during each in-sample period.
The naïve 1 / m and the MCWP portfolios are continuously updated as soon as any stock on the S&P 500 index listed as of June 2015 declares its initial public offering (IPO).
Winsorizing data at \(\beta \) percentile means replacing the extreme values of a data set with their corresponding \(\beta \) percentile values to limit the effect of the extreme values on the test. We take \(\beta = 0.90\) in our study.
Given two strategies \(s_1\) and \(s_2\), with \(\mu _{s_1}\), \(\mu _{s_2}\) as their sample means and \(\sigma ({s_1 - s_2})\) as the standard deviation of the difference of two strategies over a sample period of size n (\(n= 66\) in our case). We evaluate the p values of the difference using the t test statistic:
$$\begin{aligned} t_{\mu } := \frac{\mu _{s_1} -\mu _{s_2}}{\sigma ({s_1 - s_2})/\sqrt{n}}. \end{aligned}$$Given two strategies \(s_1\) and \(s_2\), with \(\mu _{s_1}\), \(\mu _{s_2}\), \(\sigma _{s_1}\), \(\sigma _{s_2}\), \(\sigma _{s_1,s_2}\) as their sample means, standard deviations, and the covariance of two strategies over a sample period n. We evaluate the p values by calculating the z test statistic:
$$\begin{aligned} z_{\mathrm{SR}} := \frac{\sigma _{s_2}\mu _{s_1} - \sigma _{s_1}\mu _{s_2}}{\sqrt{\Upsilon }} \end{aligned}$$with
$$\begin{aligned} \Upsilon = \frac{1}{n}\left( 2\sigma _{s_1}^2\sigma _{s_2}^2 - 2\sigma _{s_1}\sigma _{s_2}\sigma _{s_1,s_2} + 0.5\mu _{s_1}^2\sigma _{s_2}^2 + 0.5\mu _{s_2}^2\sigma _{s_1}^2 - \frac{\mu _{s_1}\mu _{s_2}}{\sigma _{s_1}\sigma _{s_2}}\sigma _{s_1,s_2}^2\right) . \end{aligned}$$Given a strategy \(s_1\) and a target portfolio \(s^*\), with \(y_1,\ldots ,y_n\) as the return series of \(s_1\) sorted from lowest to highest, \( {\widehat{\mathrm{CVaR}}_{\alpha }}, {\widehat{\mathrm{VaR}}_{\alpha }}\) as their sample \( {\mathrm{CVaR}_{\alpha }}\) and \( {\mathrm{CVaR}_{\alpha }}\) values over sample period n, we evaluate the p values by calculating the z test statistic:
$$\begin{aligned} {z_{\mathrm{CVaR}_{\alpha }}} := \frac{ {\sqrt{n(1-\alpha )} (c - \widehat{\mathrm{CVaR}}_{\alpha })}}{\sqrt{\displaystyle \sum _{i=n\alpha + 1}^{n} (y_i - {\widehat{\mathrm{CVaR}}_{\alpha }})^2/\big (n(1-\alpha )\big ) + \alpha \big ({\widehat{\mathrm{CVaR}}_{\alpha }} - {\widehat{\mathrm{VaR}}_{\alpha }}\big )^2}},\quad \mathrm{where}\quad {\widehat{\mathrm{VaR}}_{\alpha }} := y_{n\alpha }, \end{aligned}$$and
$$\begin{aligned} {\widehat{\mathrm{CVaR}}_{\alpha }} := \frac{1}{n(1-\alpha )}\displaystyle \sum _{i=n\alpha + 1}^{n}y_i. \end{aligned}$$Given two strategies \(s_1\) and \(s_2\), with \(\bar{\sigma }^2_{s_1}\), \(\bar{\sigma }^2_{s_2}\) as their sample variances over a sample period of size n (\(n= 66\) in our case), we evaluate the p values using F test statistic: \(F_{\sigma ^2} := \displaystyle \frac{\bar{\sigma }^2_{s_2}}{\bar{\sigma }^2_{s_1}}\) (Table 4).
References
Artzner P, Delbaen F, Eber J, Heath D (1999) Coherent measures of risk. Math Finance 9(3):203–228
Avouyi-Dovi S, Morin S, Neto D (2004) Optimal asset allocation with omega function. In: Technical report, Banque de France
Ben-Tal A, Ghaoui LE, Nemirovski A (2009) Robust optimization. Princeton series in applied mathematics. Princeton University Press, Princeton
Ben-Tal A, Nemirovski A (1999) Robust solutions of uncertain linear programs. Oper Res Lett 25(1):1–13
Ben-Tal A, Nemirovski A (2000) Robust solutions of linear programming problems contaminated with uncertain data. Math Program 88(3):411–424
Bertsimas D, Brown D, Caramanis C (2011) Theory and applications of robust optimization. SIAM Rev 53(3):464–501
Bertsimas D, Sim M (2004) The price of robustness. Oper Res 52(1):35–53
Beyer HG, Sendhoff B (2007) Robust optimization a comprehensive survey. Comput Methods Appl Mech Eng 196(33–34):3190–3218
Biglova A, Ortobelli S, Rachev ST, Stoyanov S (2004) Different approaches to risk estimation in portfolio theory. J Portf Manag 31:103–112
Birge JR, Louveaux FV (2011) Introduction to stochastic programming. Springer series in operations research and financial engineering. Springer, New York
Calafiore G, Ghaoui LE (2014) Optimization models. Cambridge University Press, Cambridge
Charnes A, Cooper W (1962) Programming with linear fractional functionals. Naval Res Logist Q 9(3–4):181–186
Chen C, Kwon RH (2012) Robust portfolio selection for index tracking. Comput Oper Res 39(4):829–837
Dembo R, Mausser H (2000) The put/call efficient frontier. Algo Res Q 3(1):13–26
DeMiguel V, Garlappi L, Uppal R (2009) Optimal versus naive diversification: how inefficient is the 1/\(n\) portfolio strategy? Rev Financ Stud 22(5):1915–1953
Dueck G, Tobias S (1990) Threshold accepting. A general purpose optimization algorithm superior to simulated annealing. J Comput Phys 90(1):161–175
Fishburn P (1977) Mean-risk analysis with risk associated with below-target returns. Am Econ Rev 67(2):116–126
Fliege J, Werner R (2014) Robust multiobjective optimization & applications in portfolio optimization. Eur J Oper Res 234(2):422–433
Ghaoui LE, Oks M, Oustry F (2003) Worst-case value-at-risk and robust portfolio optimization: a conic programming approach. Oper Res 51(4):543–556
Gilli M, Schumann E, Tollo GD, Cabej G (2011) Constructing 130/30-portfolios with the omega ratio. J Asset Manag 12(2):94–108
Glover F, Laguna M (1997) Tabu search. Kluwer Academic, Norwell
Gregoriou G, Gueyie J (2003) Risk-adjusted performance of funds of hedge funds using a modified Sharpe ratio. J Wealth Manag 6(3):77–83
Guastaroba G, Mansini R, Ogryczak W, Speranza MG (2016) Linear programming models based on omega ratio for the enhanced index tracking problem. Eur J Oper Res 251(3):938–956
Huyer W, Neumaier A (1999) Global optimization by multilevel coordinate search. J Glob Optim 14(4):331–355
Jensen MC (1967) The performance of mutual funds in the period 1945–1964. J Finance 23(2):389–416
Jobson J, Korkie B (1981) Performance hypothesis testing with the Sharpe and Treynor measures. J Finance 36(4):889–908
Kane SJ, Bartholomew-Biggs MC, Cross M, Dewar M (2009) Optimizing omega. J Glob Optim 45(1):153–167
Kapos M, Christofides N, Rustem B (2014) Worst-case robust omega ratio. Eur J Oper Res 234(2):499–507
Kapos M, Zymler S, Christotide N, Rustem B (2014) Optimizing the omega ratio using linear programming. J Comput Finance 17:49–57
Keating C, Shadwick W (2002) A universal performance measure. In: Technical report, The Finance Development Centre
Kirilyuk V (2013) Maximizing the omega ratio by two linear programming problems. Cybern Syst Anal 49(5):699–705
Konno H, Yamazaki H (1991) Mean-absolute deviation portfolio optimization model and its application to Tokyo stock market. Manag Sci 37(5):519–531
Linsmeier TJ, Pearson ND (1996) Risk measurement: an introduction to value-at-risk. In: Technical report 96-04, OFOR, University of Illinois, Urbana-Champaign
Lobo MS, Vandenberghe L, Boyd S, Lebret H (1998) Applications of second-order cone programming. Linear Algebra Appl 284(1–3):193–228
Mansini R, Ogryczak W, Speranza MG (2003) On lp solvable models for portfolio selection. Informatica 14(1):37–62
Markowitz H (1959) Portfolio selection: efficient diversification of investments. Wiley, New York
Markowitz HM (1952) Portfolio selection. J Finance 7(1):77–91
Martin RD, Rachev ST, Siboulet F (2003) Phi-alpha optimal portfolios and extreme risk management. In: Wilmott magazine of finance, pp 70–83
Mausser H, Saunders D, Seco L (2006) Optimizing omega. Risk Mag 11:88–92
Moon Y, Yao T (2011) A robust mean absolute deviation model for portfolio optimization. Comput Oper Res 38(9):1251–1258
Newbould GD, Poon PS (1996) Portfolio risk, portfolio performance, and individual investor. J Invest 5(2):72–78
Ogryczak W, Ruszczynski A (1999) From stochastic dominance to mean-risk models: semideviations as risk measures. Eur J Oper Res 116(1):33–50
Rachev ST, Ortobelli S, Stoyanov S, Fabozzi F, Biglova A (2008) Desirable properties of an ideal risk measure in portfolio theory. Int J Theor Appl Finance 11(1):19–54
Reeves C (1993) Modern heuristic techniques for combinatorial problems. Wiley, New York
Rockafellar RT, Uryasev S (2000) Optimization of conditional value-at-risk. J Risk 2:21–42
Rockafellar RT, Uryasev S (2002) Conditional value-at-risk for general loss distributions. J Bank Finance 26(7):1443–1471
Roman D, Mitra G (2009) Portfolio selection models: a review and new directions. Wilmott J 1(2):69–85
Ruszczynski AP, Shapiro A (2003) Stochastic programming. In: Handbooks in operations research and management science, vol 10. Elsevier, New York
Sharma A, Mehra A (2015) Extended omega ratio optimization for risk-averse investors. Int Transact Oper Res (forthcoming)
Sharpe WF (1966) Mutual fund performance. J Bus 39:119–138
Sharpe WF (1994) The Sharpe ratio. J Portf Manag 21(1):49–58
Shawky HA, Smith DM (2005) Optimal number of stock holdings in mutual fund portfolios based on market performance. Financ Rev 40(4):481–495
Sortino FA, Price LN (1994) Performance measurement in a downside risk framework. J Invest 3(3):59–64
Stoyanov S, Rachev ST, Fabozzi F (2007) Optimal financial portfolios. Appl Math Finance 14(5):401–436
Treynor JL (1965) How to rate management of investment funds. Harv Bus Rev 43(1):63–75
Vekas P (2015) An asymptotic test for the conditional value-at-risk with applications in capital adequacy. In: Presented to the actuaries institute ASTIN, AFIR/ERM and IACA Colloquia
Young MR (1998) A minimax portfolio selection rule with linear programming solution. Manag Sci 44(5):673–683
Zhu S, Fukushima M (2009) Worst-case conditional value-at-risk with application to robust portfolio management. Oper Res 57(5):1155–1168
Acknowledgments
We thank two anonymous OR Spectrum reviewers for their helpful advice and constructive comments on an earlier draft.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix A: Worst Omega ratio in continuous case under mixed uncertainty set
Define
Applying the Charnes and Cooper (1962) transformation, with \(\gamma \, > \, 0\) as a homogenization variable, we have,
where \(G^k_1 = \displaystyle \int _{\ell }(L\gamma - \ell _{\widetilde{x}})^+p^k(\ell )\,\mathrm{d}\ell , \; G^k_2 = \displaystyle \int _{\ell }(\ell _{\widetilde{x}} - L\gamma )^+p^k(\ell )\,\mathrm{d}\ell \).
Appendix B: Mixed uncertainty set
The mixed uncertainty set for the distribution of benchmark market loss \(\ell _z\) is as follows:
where \(q^k\) is the kth likelihood density function of portfolio loss \(\hat{\ell }_z\). The worst case analysis of \( {P_{\mathrm{CVaR}_{\alpha }(\hat{\ell }_z)}}\) under the mixed uncertainty set for continuous case is already discussed in Zhu and Fukushima (2009). Here we re-state its discrete version for reader comprehension. Let \(\hat{T}^k\) be the finite number of scenarios of \(\hat{\ell }_z\) (using sampling techniques) with the kth, \(k=1,\ldots ,\hat{s}\) likelihood probability vector \(q^k = ((q^k_1,\ldots ,q^k_{\hat{T}^k})^t; \; (q^k)^te =1,\; q^k_i \ge 0,\, \forall \,i=1,\ldots ,\hat{T}^k)\). Then following Zhu and Fukushima (2009), the \( {P_{\mathrm{CVaR}_{\alpha }(\hat{\ell }_z)}}\) model under \(\mathcal {Q}_M\) as follows:
where \(\hat{B}^k = [\hat{\ell }^k_{ij}]_{\hat{T}^k \times m}\) is the loss matrix of portfolio z corresponding to the likelihood probability function \(q^k\).
Therefore, the dual of \( {P_{\mathrm{MCVaR}_{\alpha }(\hat{\ell }_z)}}\) is derived as follows:
where \(v_k,\, k=1,\ldots ,\hat{s}\), is the kth component of vector v.
Appendix C: Box uncertainty set
Let \(\mathcal {Q}_B = \{q = q^0 + \hat{\pi };\; \hat{\pi }^te = 0,\; \underline{\hat{\pi }} \le \hat{\pi } \le \bar{\hat{\pi }} \}\) be a box uncertainty set for the distribution of benchmark market loss \(\hat{\ell _z}\), then \({P_{\mathrm{CVaR}_{\alpha }(\hat{\ell }_z)}}\) under \(\mathcal {Q}_B\) is given according to Zhu and Fukushima (2009):
where \(\hat{B} = [\hat{\ell }_{ij}]_{\hat{T} \times m}\) is the loss matrix of portfolio z. The dual of \({P_{\mathrm{BCVaR}_{\alpha }(\hat{\ell }_z)}}\) is as follows:
Appendix D: Ellipsoidal uncertainty set
For fixed values of \(\theta , \widetilde{u}, \widetilde{d}\), here, we first derive the dual of \(P_{10}\) in the following steps: The Lagrange of \(P_{10}\) with Lagrange multipliers \(\lambda _1,\,\lambda _2,\,\lambda _3,\,\lambda _4,\,\lambda _5\), is given as follows:
Using the minimax representation of the primal problem \(P_{11}\) as
the inner problem of the latter one is affine in \(\pi \) and can be solved by taking its derivative with respect to \(\pi \) which leads to the following dual constraint derived from Eq. (24):
Therefore, the dual of \(P_{11}\) is given as follows:
\(P_{12}\) is an SOCP problem.
Dual of worst case of \( {\mathrm{CVaR}_{\alpha }}\) model under ellipsoidal case
For the ellipsoidal set \(\mathcal {Q}_E = \{ q = q^0 + \hat{A}\hat{\pi };\; e^t\hat{A}\hat{\pi } = 0,\; q^0 + \hat{A}\hat{\pi } \ge 0,\; ||\hat{\pi }||_2 \le 1 \}\) of portfolio z, the worst case of \({P_{\mathrm{CVaR}_{\alpha }(\ell _z)}}\) is as follows (Zhu and Fukushima 2009):
\({P_{\mathrm{ECVaR}_{\alpha }(\hat{\ell }_z)}}\) is an SOCP problem. Analogously to the dual derivation \(P_{12}\) from \(P_{11}\), we can also obtain the dual of \({P_{\mathrm{ECVaR}_{\alpha }(\hat{\ell }_z)}}\) as follows:
Rights and permissions
About this article
Cite this article
Sharma, A., Utz, S. & Mehra, A. Omega-CVaR portfolio optimization and its worst case analysis. OR Spectrum 39, 505–539 (2017). https://doi.org/10.1007/s00291-016-0462-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00291-016-0462-y