pith. sign in

arxiv: 2503.08272 · v3 · pith:WGQQML3Onew · submitted 2025-03-11 · 💱 q-fin.PM · math.OC

Dynamically optimal portfolios for monotone mean--variance preferences

Ale\v{s} \v{C}ern\'y , Johannes Ruf , Martin Schweizer This is my paper

Pith reviewed 2026-05-23 00:57 UTC · model grok-4.3

classification 💱 q-fin.PM math.OC
keywords monotone mean-variance utilitydynamic portfolio choicemonotone Sharpe ratioindependent returnsmean-variance efficiencyportfolio optimization
0
0 comments X

The pith

Monotone mean-variance utility admits a full dynamic portfolio characterization in independent-return models under assumptions weaker than an equivalent martingale measure.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes a complete characterization of optimal dynamic portfolios for monotone mean-variance utility. This is carried out in asset price models with independent returns, using assumptions weaker than the existence of an equivalent martingale measure and without any restrictions on return moments. The maximal MMV utility is expressed through the monotone Sharpe ratio, where the global squared ratio equals the nominal yield obtained by continuously compounding the maximal local squared ratio. The work also supplies necessary and sufficient conditions under which mean-variance efficient portfolios remain MMV efficient.

Core claim

This paper provides, for the first time, a complete characterization of optimal dynamic portfolio choice for the MMV utility in asset price models with independent returns. The task is performed under minimal assumptions, weaker than the existence of an equivalent martingale measure and with no restrictions on the moments of asset returns. The maximal MMV utility is interpreted in terms of the monotone Sharpe ratio and the global squared MSR arises as the nominal yield from continuously compounding at the rate equal to the maximal local squared MSR. Simple necessary and sufficient conditions are given for mean-variance efficient portfolios to be MMV efficient.

What carries the argument

The monotone Sharpe ratio, which quantifies maximal MMV utility and enables separation of the global problem into local optimization steps via the independent-returns assumption.

If this is right

  • The global squared monotone Sharpe ratio equals the compounded yield of successive local maximal squared ratios.
  • Mean-variance efficient portfolios satisfy explicit necessary and sufficient conditions to remain efficient under the monotone criterion.
  • Optimal MMV portfolios exist and can be constructed without invoking an equivalent martingale measure or finite-moment assumptions.
  • The maximal MMV utility admits a direct interpretation as the monotone Sharpe ratio of the optimal strategy.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The independence assumption may allow similar separation results for other time-separable preference classes that admit a local-global decomposition.
  • The compounding representation suggests that long-horizon MMV problems can be solved by iterating short-horizon local problems in discrete time.
  • The absence of moment restrictions implies the characterization remains valid in models with heavy-tailed returns, provided independence holds.

Load-bearing premise

Asset returns are independent across time periods.

What would settle it

A counter-example in which returns lack independence yet the derived local-to-global compounding relation for the squared monotone Sharpe ratio still holds.

Figures

Figures reproduced from arXiv: 2503.08272 by Ale\v{s} \v{C}ern\'y, Johannes Ruf, Martin Schweizer.

Figure 1
Figure 1. Figure 1: Quadratic utility and its monotonization. where the supremum is taken over Y ≥ 0 such that W − Y ∈ L 2 . The monotone mean– variance utility Vmmv sets aside a non-negative amount Y to make the mean–variance utility of the remaining wealth W − Y as high as possible. Observe that thanks to (1.2) the domain of the MMV utility extends beyond L 2 , which will later allow us to consider returns that are not squa… view at source ↗
read the original abstract

Monotone mean-variance (MMV) utility is the minimal modification of the classical Markowitz utility that respects rational ordering of investment opportunities. This paper provides, for the first time, a complete characterization of optimal dynamic portfolio choice for the MMV utility in asset price models with independent returns. The task is performed under minimal assumptions, weaker than the existence of an equivalent martingale measure and with no restrictions on the moments of asset returns. We interpret the maximal MMV utility in terms of the monotone Sharpe ratio (MSR) and show that the global squared MSR arises as the nominal yield from continuously compounding at the rate equal to the maximal local squared MSR. The paper gives simple necessary and sufficient conditions for mean-variance (MV) efficient portfolios to be MMV efficient. Several illustrative examples contrasting the MV and MMV criteria are provided.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 1 minor

Summary. The paper claims to provide the first complete characterization of optimal dynamic portfolio choice for monotone mean-variance (MMV) utility in asset price models with independent returns. This is achieved under minimal assumptions weaker than the existence of an equivalent martingale measure and with no restrictions on the moments of asset returns. The maximal MMV utility is interpreted in terms of the monotone Sharpe ratio (MSR), and the global squared MSR is shown to arise as the nominal yield from continuously compounding at the maximal local squared MSR rate. Simple necessary and sufficient conditions are given for mean-variance efficient portfolios to also be MMV efficient, supported by illustrative examples contrasting the two criteria.

Significance. If the central derivations and proofs hold, the result is significant for extending dynamic portfolio theory to MMV preferences under notably weak assumptions. The separation into local and global problems, together with the explicit compounding interpretation of the squared MSR, provides a clean link between periods that relies only on the stated independence of returns. The necessary and sufficient conditions for MV-to-MMV efficiency and the absence of moment restrictions or EMM strengthen the applicability relative to classical MV analysis.

minor comments (1)
  1. The abstract is dense; expanding the statement of the independence assumption and its precise role in the separation result would improve readability for readers outside the immediate subfield.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their thorough reading and positive evaluation of the manuscript. We are pleased that the referee recognizes the significance of the results, particularly the complete characterization under minimal assumptions, the link to the monotone Sharpe ratio, and the conditions distinguishing MV and MMV efficiency.

Circularity Check

0 steps flagged

No significant circularity

full rationale

The derivation separates local and global optimization problems using the explicit assumption of independent returns, then interprets the global squared MSR as the compounded yield from the maximal local squared MSR. These relations are obtained from the model dynamics and the definition of MMV utility rather than by re-labeling fitted quantities or reducing to self-citations. The central characterization holds under the stated weaker-than-EMM assumptions without any load-bearing step shown to be definitionally equivalent to its inputs. The paper is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Based solely on the abstract, the central claim rests on the domain assumption of independent returns and standard mathematical background for stochastic processes; no free parameters or invented entities are mentioned.

axioms (1)
  • domain assumption Asset returns are independent
    The complete characterization and the local-to-global compounding result are stated for asset price models with independent returns.

pith-pipeline@v0.9.0 · 5677 in / 1245 out tokens · 63108 ms · 2026-05-23T00:57:03.838773+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

55 extracted references · 55 canonical work pages

  1. [1]

    Bellini and M

    F. Bellini and M. Frittelli. On the existence of minimax martingale measures.Math. Finance, 12(1):1–21, 2002. MR1883783

    work page 2002

  2. [2]

    Bender and C

    C. Bender and C. R. Niethammer. Onq-optimal martingale measures in exponential L´ evy models.Finance Stoch., 12(3):381–410, 2008. MR2410843

    work page 2008

  3. [3]

    Biagini and A

    S. Biagini and A. ˇCern´ y. Admissible strategies in semimartingale portfolio selection.SIAM J. Control Optim., 49(1):42–72, 2011. MR2765656

    work page 2011

  4. [4]

    Biagini and A

    S. Biagini and A. ˇCern´ y. Convex duality and Orlicz spaces in expected utility maximization. Math. Finance, 30(1):85–127, 2020. MR4067071

    work page 2020

  5. [5]

    Black and A

    F. Black and A. F. Perold. Theory of constant proportion portfolio insurance.J. Econom. Dynam. Control, 16(3):403–426, 1992

    work page 1992

  6. [6]

    Cawston and L

    S. Cawston and L. Vostrikova. Anf-divergence approach for optimal portfolios in exponen- tial L´ evy models. In Y. Kabanov, M. Rutkowski, and T. Zariphopoulou, editors,Inspired by Finance, pages 83–101. Springer, Cham, 2014. The Musiela Festschrift. MR3204213

    work page 2014

  7. [7]

    ˇCern´ y

    A. ˇCern´ y. Generalized Sharpe ratios and asset pricing in incomplete markets.European Finance Review, 7(2):191–233, 2003

    work page 2003

  8. [8]

    ˇCern´ y

    A. ˇCern´ y. Semimartingale theory of monotone mean–variance portfolio allocation.Math. Finance, 30(3):1168–1178, 2020. MR4116461

    work page 2020

  9. [9]

    ˇCern´ y

    A. ˇCern´ y. The Hansen ratio in mean–variance portfolio theory. Available from arXiv:2007.15980, 2020

    work page arXiv 2007

  10. [10]

    ˇCern´ y and J

    A. ˇCern´ y and J. Kallsen. On the structure of general mean–variance hedging strategies. Ann. Probab., 35(4):1479–1531, 2007. MR2330978

    work page 2007

  11. [11]

    ˇCern´ y and J

    A. ˇCern´ y and J. Ruf. Pure-jump semimartingales.Bernoulli, 27(4):2624–2648, 2021. MR4303898

    work page 2021

  12. [12]

    ˇCern´ y and J

    A. ˇCern´ y and J. Ruf. Simplified stochastic calculus with applications in Economics and Finance.European J. Oper. Res., 293(2):547–560, 2021. MR4241583

    work page 2021

  13. [13]

    ˇCern´ y and J

    A. ˇCern´ y and J. Ruf. Simplified stochastic calculus via semimartingale representations. Electron. J. Probab., 27:1–32, 2022. Paper No. 3. MR4362774

    work page 2022

  14. [14]

    ˇCern´ y and J

    A. ˇCern´ y and J. Ruf. Simplified calculus for semimartingales: Multiplicative compensators and changes of measure.Stochastic Process. Appl., 161:572–602, 2023. MR4585465

    work page 2023

  15. [15]

    ˇCern´ y, F

    A. ˇCern´ y, F. Maccheroni, M. Marinacci, and A. Rustichini. On the computation of optimal monotone mean–variance portfolios via truncated quadratic utility.J. Math. Econom., 48 (6):386–395, 2012. MR2988098

    work page 2012

  16. [16]

    X. Cui, D. Li, S. Wang, and S. Zhu. Better than dynamic mean–variance: Time inconsis- tency and free cash flow stream.Math. Finance, 22(2):346–378, 2012. MR2897388

    work page 2012

  17. [17]

    Dol´ eans-Dade

    C. Dol´ eans-Dade. Quelques applications de la formule de changement de variables pour les semimartingales.Z. Wahrsch. verw. Gebiete, 16:181–194, 1970. MR283883

    work page 1970

  18. [18]

    Eberlein and J

    E. Eberlein and J. Jacod. On the range of options prices.Finance Stoch., 1(2):131–140, 1997

    work page 1997

  19. [19]

    M. ´Emery. Stabilit´ e des solutions des ´ equations diff´ erentielles stochastiques application aux int´ egrales multiplicatives stochastiques.Z. Wahrsch. verw. Gebiete, 41(3):241–262, 1978. MR0464400

    work page 1978

  20. [20]

    Esche and M

    F. Esche and M. Schweizer. Minimal entropy preserves the L´ evy property: how and why. Stochastic Process. Appl., 115(2):299–327, 2005. MR2111196 37

    work page 2005

  21. [21]

    Filipovi´ c and M

    D. Filipovi´ c and M. Kupper. Monotone and cash-invariant convex functions and hulls. Insurance Math. Econom., 41(1):1–16, 2007. MR2324561

    work page 2007

  22. [22]

    Fujiwara

    T. Fujiwara. The minimal entropy martingale measures for exponential additive processes revisited.J. Math-for-Ind., 2B:115–125, 2010. MR2734483

    work page 2010

  23. [23]

    Fujiwara and Y

    T. Fujiwara and Y. Miyahara. The minimal entropy martingale measures for geometric L´ evy processes.Finance Stoch., 7(4):509–531, 2003. MR2014248

    work page 2003

  24. [24]

    Goll and L

    T. Goll and L. R¨ uschendorf. Minimax and minimal distance martingale measures and their relationship to portfolio optimization.Finance Stoch., 5(4):557–581, 2001. MR1862002

    work page 2001

  25. [25]

    S. Hodges. A generalization of the Sharpe Ratio and its application to valuation bounds and risk measures. Available from warwick.ac.uk. Accessed 27/12/2024. FORC Preprint 98/88, University of Warwick, 1998

    work page 2024

  26. [26]

    Y. Hu, X. Shi, and Z. Q. Xu. Constrained monotone mean–variance problem with random coefficients.SIAM J. Financial Math., 14(3):838–854, 2023. MR4621927

    work page 2023

  27. [27]

    Jacod and A

    J. Jacod and A. N. Shiryaev.Limit Theorems for Stochastic Processes, volume 288 of Comprehensive Studies in Mathematics. Springer, Berlin, 2nd edition, 2003. MR1943877

    work page 2003

  28. [28]

    Jakub˙ enas

    P. Jakub˙ enas. On option pricing in certain incomplete markets. In A. N. Shiryaev and E. F. Mishchenko, editors,Stochastic Financial Mathematics, volume 237 ofTr. Mat. Inst. Steklova, pages 123–142. Nauka, Moscow, 2002. MR1976510

    work page 2002

  29. [29]

    Jeanblanc, S

    M. Jeanblanc, S. Kl¨ oppel, and Y. Miyahara. Minimalf q-martingale measures of exponen- tial L´ evy processes.Ann. Appl. Probab., 17(5-6):1615–1638, 2007. MR2358636

    work page 2007

  30. [30]

    J. Kallsen. A utility maximization approach to hedging in incomplete markets.Math. Methods Oper. Res., 50(2):321–338, 1999. MR1732402

    work page 1999

  31. [31]

    J. Kallsen. Optimal portfolios for exponential L´ evy processes.Math. Methods Oper. Res., 51(3):357–374, 2000. MR1778648

    work page 2000

  32. [32]

    Kallsen.σ-localization andσ-martingales.Theory Probab

    J. Kallsen.σ-localization andσ-martingales.Theory Probab. Appl., 48(1):152–163, 2004. MR2013413

    work page 2004

  33. [33]

    Kallsen and J

    J. Kallsen and J. Muhle-Karbe. Utility maximization in models with conditionally inde- pendent increments.Ann. Appl. Probab., 20(6):2162–2177, 2010. MR2759731

    work page 2010

  34. [34]

    Karatzas and C

    I. Karatzas and C. Kardaras. The num´ eraire portfolio in semimartingale financial models. Finance Stoch., 11(4):447–493, 2007. MR2335830

    work page 2007

  35. [35]

    Kardaras

    C. Kardaras. No-free-lunch equivalences for exponential L´ evy models under convex con- straints on investment.Math. Finance, 19(2):161–187, 2009. MR2509267

    work page 2009

  36. [36]

    B. Li, J. Guo, and L. Tian. Optimal investment and reinsurance policies for the Cram´ er– Lundberg risk model under monotone mean–variance preference.Internat. J. Control, 97 (6):1296–1310, 2024. MR4742973

    work page 2024

  37. [37]

    Y. Li, Z. Liang, and S. Pang. Continuous-time monotone mean–variance portfolio selection. Available from arXiv:2211.12168v3, 2023

    work page arXiv 2023

  38. [38]

    Y. Li, Z. Liang, and S. Pang. Comparison between mean–variance and monotone mean– variance preferences under jump diffusion and stochastic factor model.Math. Oper. Res., 50(3):2405–2432, 2025. MR4949476

    work page 2025

  39. [39]

    Y. Li, Z. Liang, and S. Pang. Comparison between mean–variance and monotone mean– variance preferences in general markets: a new perspective.Oper. Res. Lett., 61:Paper No. 107298, 7, 2025. MR4896371

    work page 2025

  40. [40]

    Maccheroni, M

    F. Maccheroni, M. Marinacci, and A. Rustichini. Ambiguity aversion, robustness, and the variational representation of preferences.Econometrica, 74(6):1447–1498, 2006. MR2268407 38

    work page 2006

  41. [41]

    Maccheroni, M

    F. Maccheroni, M. Marinacci, A. Rustichini, and M. Taboga. Portfolio selection with monotone mean–variance preferences.Math. Finance, 19(3):487–521, 2009. MR2536871

    work page 2009

  42. [42]

    M. Nutz. Power utility maximization in constrained exponential L´ evy models.Math. Finance, 22(4):690–709, 2012. MR2968281

    work page 2012

  43. [43]

    Pennanen and A.-P

    T. Pennanen and A.-P. Perkki¨ o.Convex Stochastic Optimization, volume 107 ofProbability Theory and Stochastic Modelling. Springer Cham, 2024

    work page 2024

  44. [44]

    P. E. Protter.Stochastic Integration and Differential Equations, volume 21 ofStochastic Modelling and Applied Probability. Springer, Berlin, 2nd edition, 2005. MR2273672

    work page 2005

  45. [45]

    R. T. Rockafellar.Convex Analysis. Princeton Mathematical Series, No. 28. Princeton University Press, Princeton, N.J., 1970. MR0274683

    work page 1970

  46. [46]

    R. T. Rockafellar and R. J.-B. Wets.Variational Analysis, volume 317 ofGrundlehren der mathematischen Wissenschaften. Springer-Verlag, Berlin, 1998. MR1491362

    work page 1998

  47. [47]

    Rudin.Real and Complex Analysis

    W. Rudin.Real and Complex Analysis. McGraw-Hill Book Co., New York, 3rd edition,

    work page

  48. [48]

    Schweizer

    M. Schweizer. Approximating random variables by stochastic integrals.Ann. Probab., 22 (3):1536–1575, 1994. MR1303653

    work page 1994

  49. [49]

    Schweizer

    M. Schweizer. Approximation pricing and the variance-optimal martingale measure.Ann. Probab., 24(1):206–236, 1996. MR1387633

    work page 1996

  50. [50]

    Shen and B

    Y. Shen and B. Zou. Short communication: Cone-constrained monotone mean–variance portfolio selection under diffusion models.SIAM J. Financial Math., 13(4):SC99–SC112,

    work page

  51. [51]

    Shi and Z

    X. Shi and Z. Q. Xu. Constrained monotone mean–variance investment-reinsurance under the Cram´ er-Lundberg model with random coefficients.Systems Control Lett., 188:Paper No. 105796, 9, 2024. MR4732938

    work page 2024

  52. [52]

    A. N. Shiryaev and A. S. Cherny. A vector stochastic integral and the fundamental theorem of asset pricing.Proc. Steklov Inst. Math., 237:6–49, 2002. MR1975582

    work page 2002

  53. [53]

    M. S. Strub and D. Li. A note on monotone mean–variance preferences for continuous processes.Oper. Res. Lett., 48(4):0–3, 2020. MR4100873

    work page 2020

  54. [54]

    Trybu la and D

    J. Trybu la and D. Zawisza. Continuous-time portfolio choice under monotone mean– variance preferences—stochastic factor case.Math. Oper. Res., 44(3):966–987, 2019. MR3996654

    work page 2019

  55. [55]

    Zhitlukhin

    M. Zhitlukhin. Monotone Sharpe ratios and related measures of investment performance. In 2017 MATRIX Annals, volume 2 ofMATRIX Book Ser., pages 637–665. Springer, Cham,

    work page 2017