pith. sign in

arxiv: 2406.12305 · v5 · submitted 2024-06-18 · 💱 q-fin.MF · math.OC· math.PR· q-fin.GN

Robust dividend policy: Equivalence of Epstein-Zin and Maenhout preferences

Kexin Chen , Kyunghyun Park , Hoi Ying Wong This is my paper

Pith reviewed 2026-05-24 00:20 UTC · model grok-4.3

classification 💱 q-fin.MF math.OCmath.PRq-fin.GN
keywords Epstein-Zin preferencesMaenhout ambiguity aversionsingular stochastic controldividend policyrobust controlHamilton-Jacobi-Bellman variational inequalitybackward stochastic differential equation
0
0 comments X

The pith

The Epstein-Zin preference for discounted dividends is equivalent to a robust dividend policy under Maenhout's ambiguity-averse preference.

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

This paper formulates Epstein-Zin preferences over dividends as a singular control utility via a backward stochastic differential equation whose aggregator integrates against the singular control. It proves the BSDE is well-posed and recovers the Epstein-Zin utility, then shows this object is mathematically identical to the dividend problem solved by a firm executive under Maenhout ambiguity aversion. The common solution is a threshold strategy on the surplus process whose level is the free boundary of an associated Hamilton-Jacobi-Bellman variational inequality. A sympathetic reader cares because the equivalence supplies a single object that simultaneously encodes investor time-risk preferences and executive robustness concerns.

Core claim

We formulate the Epstein-Zin preference for discounted dividends received by an investor as an Epstein-Zin singular control utility. We introduce a backward stochastic differential equation with an aggregator integrated with respect to a singular control, prove its well-posedness, and show that it coincides with the Epstein-Zin singular control utility. We then establish that this formulation is equivalent to a robust dividend policy chosen by the firm's executive under Maenhout's ambiguity-averse preference. In particular, the robust dividend policy takes the form of a threshold strategy on the firm's surplus process, where the threshold level is characterized as the free boundary of a HJB-

What carries the argument

The backward stochastic differential equation with aggregator integrated with respect to a singular control, which serves as the common mathematical object linking Epstein-Zin utility to Maenhout robust control.

If this is right

  • The robust dividend policy is a threshold strategy on the firm's surplus process.
  • The threshold level equals the free boundary of the associated Hamilton-Jacobi-Bellman variational inequality.
  • Investors can match their preferences to firms by inspecting observed dividend policies and financial statements.
  • Executives can use dividend thresholds to signal the degree of ambiguity aversion attached to earnings projections.

Where Pith is reading between the lines

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

  • Valuation models built on either preference specification can be interchanged without changing the implied dividend thresholds.
  • The same threshold mechanism may apply to other singular-control problems such as capital structure or inventory decisions.
  • Observed dividend thresholds in market data could be used to back out implied ambiguity-aversion parameters.

Load-bearing premise

The backward stochastic differential equation with aggregator integrated with respect to a singular control is well-posed and coincides with the Epstein-Zin singular control utility.

What would settle it

A concrete counter-example in which the BSDE solution differs from the Epstein-Zin singular control utility value, or in which the threshold strategy fails to solve the Maenhout robust-control problem.

read the original abstract

In a continuous-time economy, this paper formulates the Epstein-Zin preference for discounted dividends received by an investor as an Epstein-Zin singular control utility. We introduce a backward stochastic differential equation with an aggregator integrated with respect to a singular control, prove its well-posedness, and show that it coincides with the Epstein-Zin singular control utility. We then establish that this formulation is equivalent to a robust dividend policy chosen by the firm's executive under the Maenhout's ambiguity-averse preference. In particular, the robust dividend policy takes the form of a threshold strategy on the firm's surplus process, where the threshold level is characterized as the free boundary of a Hamilton-Jacobi-Bellman variational inequality. Therefore, dividend-caring investors can choose firms that match their preferences by examining stock's dividend policies and financial statements, whereas executives can make use of dividend to signal their confidence, in the form of ambiguity aversion, on realizing the earnings implied by their financial statements.

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

1 major / 0 minor

Summary. The paper formulates Epstein-Zin preferences over discounted dividends as a singular-control utility, introduces a BSDE whose aggregator is integrated against the singular dividend process, proves well-posedness of this BSDE and its equivalence to the Epstein-Zin utility, and then shows that the resulting problem is dual to a Maenhout ambiguity-averse robust-control problem whose optimal policy is a threshold strategy on the surplus process, with the threshold identified as the free boundary of an HJB variational inequality.

Significance. If the BSDE well-posedness and duality arguments hold, the paper supplies a direct preference-based justification for threshold dividend policies under ambiguity aversion, linking investor Epstein-Zin parameters to observable corporate payout rules and offering a signaling interpretation for executives. This is a non-trivial contribution at the intersection of recursive utility theory and singular stochastic control in corporate finance.

major comments (1)
  1. [Abstract / BSDE section] Abstract (paragraph introducing the BSDE) and the subsequent well-posedness claim: the existence/uniqueness argument for the BSDE with aggregator integrated against a singular control must explicitly treat the case in which the dividend measure has atoms (lump-sum payouts). Standard BSDE theory relies on Lipschitz or monotonicity conditions that are typically stated for continuous or absolutely continuous integrators; an atom produces a jump that can violate the contraction or comparison principle used to identify the solution with the Epstein-Zin singular-control utility. Without a separate verification or approximation argument that survives atoms, the claimed coincidence and the subsequent HJB-VI characterization of the threshold are not justified.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading and constructive feedback. The paper's contribution at the intersection of recursive utility and singular control is acknowledged, and we address the sole major comment below.

read point-by-point responses

  1. Referee: [Abstract / BSDE section] Abstract (paragraph introducing the BSDE) and the subsequent well-posedness claim: the existence/uniqueness argument for the BSDE with aggregator integrated against a singular control must explicitly treat the case in which the dividend measure has atoms (lump-sum payouts). Standard BSDE theory relies on Lipschitz or monotonicity conditions that are typically stated for continuous or absolutely continuous integrators; an atom produces a jump that can violate the contraction or comparison principle used to identify the solution with the Epstein-Zin singular-control utility. Without a separate verification or approximation argument that survives atoms, the claimed coincidence and the subsequent HJB-VI characterization of the threshold are not justified.

    Authors: We agree that an explicit argument for atoms is required to close the well-posedness claim. The manuscript's comparison principle is formulated for general finite-variation integrators, but the presentation does not isolate the atomic case. In the revision we will insert a short subsection that (i) approximates any singular dividend measure by a sequence of absolutely continuous controls, (ii) passes to the limit using the monotonicity condition already established, and (iii) verifies that the limiting process satisfies the BSDE pathwise even when atoms are present. This will also confirm that the Epstein-Zin utility identification and the subsequent HJB-VI threshold characterization remain valid under lump-sum payouts. revision: yes

Circularity Check

0 steps flagged

No circularity: equivalence derived from explicit BSDE construction and HJB analysis

full rationale

The paper defines an Epstein-Zin singular-control utility, introduces a BSDE whose aggregator is integrated against the singular dividend control, proves well-posedness of that BSDE, shows the BSDE solution coincides with the utility, and then proves equivalence of the resulting value function to the Maenhout robust-control problem via a free-boundary HJB variational inequality. None of these steps reduces by definition or by self-citation to the target equivalence; each is an independent analytic claim whose validity can be checked against the stated assumptions on the surplus process and the aggregator. No fitted parameters are relabeled as predictions, and no uniqueness theorem is imported from the authors' prior work to force the result.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The paper rests on standard existence theory for BSDEs with singular controls and on the usual domain assumptions of continuous-time stochastic control; no new free parameters or invented entities are introduced in the abstract.

axioms (2)
  • domain assumption Well-posedness of the backward stochastic differential equation with aggregator integrated w.r.t. singular control
    Invoked to equate the BSDE solution with the Epstein-Zin utility (abstract).
  • domain assumption Standard continuous-time economy with surplus process and singular dividend control
    Background modeling assumption for both preference formulations.

pith-pipeline@v0.9.0 · 5712 in / 1442 out tokens · 20662 ms · 2026-05-24T00:20:20.143451+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

70 extracted references · 70 canonical work pages

  1. [1]

    L. H. Alvarez and J. Virtanen. A class of solvable stochas tic dividend optimization problems: on the general impact of flexibility on valuation. Econ. Theory, pages 373–398, 2006

    work page 2006

  2. [2]

    E. W. Anderson, L. P. Hansen, and T. J. Sargent. A quartet o f semigroups for model spec- ification, robustness, prices of risk, and model detection. J. Eur. Econ. Assoc. , 1(1):68–123, 2003

    work page 2003

  3. [3]

    Aurand and Y.-J

    J. Aurand and Y.-J. Huang. Epstein-Zin utility maximiza tion on a random horizon. Math. Finance, 33(4):1370–1411, 2023

    work page 2023

  4. [4]

    P. Bank. Optimal control under a dynamic fuel constraint . SIAM J. Control Optim. , 44(4):1529– 1541, 2005

    work page 2005

  5. [5]

    Bank and N

    P. Bank and N. El Karoui. A stochastic representation the orem with applications to optimization and obstacle problems. Ann. Probab., 32(1B):1030–1067, 2004

    work page 2004

  6. [6]

    Bank and H

    P. Bank and H. Kauppila. Convex duality for stochastic si ngular control problems. Ann. Appl. Probab., pages 485–516, 2017

    work page 2017

  7. [7]

    Bank and F

    P. Bank and F. Riedel. Optimal consumption choice with in tertemporal substitution. Ann. Appl. Probab., 11(3):750–788, 2001

    work page 2001

  8. [8]

    Bayraktar and Y.-J

    E. Bayraktar and Y.-J. Huang. On the multidimensional co ntroller-and-stopper games. SIAM J. Control Optim. , 51(2):1263–1297, 2013

    work page 2013

  9. [9]

    Bayraktar and S

    E. Bayraktar and S. Yao. Optimal stopping for non-linear expectations—Part I. Stochastic Process. Appl., 121(2):185–211, 2011

    work page 2011

  10. [10]

    Bayraktar and S

    E. Bayraktar and S. Yao. Optimal stopping for non-linea r expectations—Part II. Stochastic Process. Appl., 121(2):212–264, 2011

    work page 2011

  11. [11]

    A. Ben-Tal. The entropic penalty approach to stochasti c programming. Math. Oper. Res. , 10(2):263–279, 1985

    work page 1985

  12. [12]

    the bird in the hand

    S. Bhattacharya. Imperfect information, dividend pol icy, and “the bird in the hand” fallacy. Rand J. Econ. , 10(1):259–270, 1979

    work page 1979

  13. [13]

    F. Black. The dividend puzzle. J. Portf. Manag. , 2(2):5–8, 1976

    work page 1976

  14. [14]

    F. Black. Noise. J. Finance , 41(3):528–543, 1986

    work page 1986

  15. [15]

    Borodin and P

    A. Borodin and P. Salminen. Handbook of Brownian Motion – Facts And Formulae . Springer Science & Business Media, 2015

    work page 2015

  16. [16]

    Branger and L

    N. Branger and L. S. Larsen. Robust portfolio choice wit h uncertainty about jump and diffusion risk. J. Bank. Finance , 37(12):5036–5047, 2013

    work page 2013

  17. [17]

    A. Brav, J. Graham, C. Harvey, and R. Michaely. Payout po licy in the 21st century. J. Financ. Econ., 77(3):483–527, 2005

    work page 2005

  18. [18]

    Briand, B

    P. Briand, B. Delyon, Y. Hu, E. Pardoux, and L. Stoica. Lp solutions of backward stochastic differential equations. Stochastic Process. Appl. , 108(1):109–129, 2003

    work page 2003

  19. [19]

    Cadenillas, S

    A. Cadenillas, S. Sarkar, and F. Zapatero. Optimal divi dend policy with mean-reverting cash reservoir. Math. Finance , 17(1):81–109, 2007

    work page 2007

  20. [20]

    Chakraborty, A

    P. Chakraborty, A. Cohen, and V. R. Young. Optimal divid ends under model uncertainty. SIAM J. Financial Math. , 14(2):497–524, 2023

    work page 2023

  21. [21]

    Cohen, A

    A. Cohen, A. Hening, and C. Sun. Optimal ergodic harvest ing under ambiguity. SIAM J. Control Optim. , 60(2):1039–1063, 2022. 42 KEXIN CHEN, KYUNGHYUN PARK, AND HOI YING WONG

    work page 2022

  22. [22]

    I. Csiszár. I-divergence geometry of probability distributions and min imization problems. Ann. Probab., pages 146–158, 1975

    work page 1975

  23. [23]

    De Angelis, S

    T. De Angelis, S. Federico, and G. Ferrari. Optimal boun dary surface for irreversible investment with stochastic costs. Math. Oper. Res. , 42(4):1135–1161, 2017

    work page 2017

  24. [24]

    Duffie and L

    D. Duffie and L. G. Epstein. Stochastic differential utili ty. Econometrica, 60(2):353–394, 1992

    work page 1992

  25. [25]

    L. G. Epstein and S. E. Zin. Substitution, risk aversion , and the temporal behavior of consump- tion and asset returns: A theoretical framework. Econometrica, 57(4):937–969, 1989

    work page 1989

  26. [26]

    G. Ferrari. On an integral equation for the free-bounda ry of stochastic, irreversible investment problems. Ann. Appl. Probab. , 25(1), 2015

    work page 2015

  27. [27]

    Ferrari, H

    G. Ferrari, H. Li, and F. Riedel. A knightian irreversib le investment problem. J. Math. Anal. Appl., 507(1):125744, 2022

    work page 2022

  28. [28]

    Ferrari, H

    G. Ferrari, H. Li, and F. Riedel. Optimal consumption wi th Hindy–Huang–Kreps preferences under nonlinear expectations. Adv. Appl. Probab. , 54(4):1222–1251, 2022

    work page 2022

  29. [29]

    W. H. Fleming and H. M. Soner. Controlled Markov processes and viscosity solutions, volume 25. Springer New York, NY, 2006

    work page 2006

  30. [30]

    C. Ham, Z. Kaplan, and R. Leary. Do dividends convey info rmation about future earnings? J. Financ. Econ. , 136(2):547–570, 2020

    work page 2020

  31. [31]

    Herdegen, D

    M. Herdegen, D. Hobson, and J. Jerome. Proper solutions for Epstein–Zin stochastic differential utility. arXiv preprint arXiv:2112.06708 , 2021

    work page arXiv 2021

  32. [32]

    Herdegen, D

    M. Herdegen, D. Hobson, and J. Jerome. The infinite-hori zon investment–consumption problem for Epstein–Zin stochastic differential utility. I: Founda tions. Finance Stoch. , 27(1):127–158, 2023

    work page 2023

  33. [33]

    Herdegen, D

    M. Herdegen, D. Hobson, and J. Jerome. The infinite-hori zon investment–consumption problem for Epstein–Zin stochastic differential utility. II: Exist ence, uniqueness and verification for ϑ ∈ (0, 1). Finance Stoch., 27(1):159–188, 2023

    work page 2023

  34. [34]

    Jacod and A

    J. Jacod and A. Shiryaev. Limit Theorems for Stochastic Processes , volume 288. Springer Science & Business Media, 2013

    work page 2013

  35. [35]

    Jeanblanc-Picqué and A

    M. Jeanblanc-Picqué and A. N. Shiryaev. Optimization o f the flow of dividends. Russian Math- ematical Surveys , 50(2):25–46, 1995

    work page 1995

  36. [36]

    X. Jin, D. Luo, and X. Zeng. Dynamic asset allocation wit h uncertain jump risks: a pathwise optimization approach. Math. Oper. Res. , 43(2):347–376, 2018

    work page 2018

  37. [37]

    John and J

    K. John and J. Williams. Dividends, dilution, and taxes : A signalling equilibrium. J. Finance , 40(4):1053–1070, 1985

    work page 1985

  38. [38]

    Karatzas and S

    I. Karatzas and S. Shreve. Brownian Motion and Stochastic Calculus . Springer-Verlag, 1991

    work page 1991

  39. [39]

    Kobylanski

    M. Kobylanski. Backward stochastic differential equat ions and partial differential equations with quadratic growth. Ann. Probab., 28(2):558 – 602, 2000

    work page 2000

  40. [40]

    Kraft, T

    H. Kraft, T. Seiferling, and F. T. Seifried. Optimal con sumption and investment with Epstein– Zin recursive utility. Finance Stoch., 21:187–226, 2017

    work page 2017

  41. [41]

    L. Kruk, J. Lehoczky, K. Ramanan, and S. Shreve. An expli cit formula for the Skorokhod map on [0, a]. Ann. Probab., 35(5):1740–1768, 2007

    work page 2007

  42. [42]

    R. J. Laeven and M. Stadje. Robust portfolio choice and i ndifference valuation. Math. Oper. Res., 39(4):1109–1141, 2014

    work page 2014

  43. [43]

    P. J. Maenhout. Robust portfolio rules and asset pricin g. Review of Financial Studies , 17(4):951–983, 2004

    work page 2004

  44. [44]

    P. J. Maenhout. Robust portfolio rules and detection-e rror probabilities for a mean-reverting risk premium. J. Econ. Theory , 128(1):136–163, 2006

    work page 2006

  45. [45]

    Matoussi and H

    A. Matoussi and H. Xing. Convex duality for Epstein–Zin stochastic differential utility. Math. Finance, 28(4):991–1019, 2018

    work page 2018

  46. [46]

    Melnyk, J

    Y. Melnyk, J. Muhle-Karbe, and F. T. Seifried. Lifetime investment and consumption with recursive preferences and small transaction costs. Math. Finance , 30(3):1135–1167, 2020

    work page 2020

  47. [47]

    Michaely, S

    R. Michaely, S. Rossi, and M. W eber. Signaling safety. J. Financ. Econ. , 139(2):405–427, 2021. 43

    work page 2021

  48. [48]

    Miller and F

    M. Miller and F. Modigliani. Dividend policy, growth, a nd the valuation of shares. The Journal of Business , 34(4):411–433, 1961

    work page 1961

  49. [49]

    Miller and K

    M. Miller and K. Rock. Dividend Policy under Asymmetric Information. J. Finance, 40(4):1031– 1051, 1985

    work page 1985

  50. [50]

    Monoyios and O

    M. Monoyios and O. Mostovyi. Stability of the Epstein–Z in problem. Math. Finance , 2024

    work page 2024

  51. [51]

    R. K. Nagle, E. B. Saff, and A. D. Snider. Fundamentals of Differential Equations And Bound- ary Value Problems . Pearson, 7 edition, 2017

    work page 2017

  52. [52]

    Nutz and J

    M. Nutz and J. Zhang. Optimal stopping under adverse non linear expectation and related games. Ann. Appl. Probab. , 25(5):2503–2534, 2015

    work page 2015

  53. [53]

    Pardoux and S

    E. Pardoux and S. Peng. Adapted solution of a backward st ochastic differential equation. Sys- tems & control letters , 14(1):55–61, 1990

    work page 1990

  54. [54]

    K. Park, K. Chen, and H. Y. W ong. Irreversible consumpti on habit under ambiguity: Singular control and optimal G-stopping time. Ann. Appl. Probab. (to appear) , 2025+

    work page 2025

  55. [55]

    Park and H

    K. Park and H. Y. W ong. Robust retirement with return amb iguity: Optimal G-stopping time in dual space. SIAM J. Control Optim. , 61(3):1009–1037, 2023

    work page 2023

  56. [56]

    K. Park, H. Y. W ong, and T. Yan. Robust retirement and lif e insurance with inflation risk and model ambiguity. Insur. Math. Econ. , 110:1–30, 2023

    work page 2023

  57. [57]

    A. Popier. Backward stochastic differential equations with random stopping time and singular final condition. Ann. Probab., pages 1071–1117, 2007

    work page 2007

  58. [58]

    P. Protter. Stochastic Integration and Differential Equations . Springer Berlin, Heidelberg, 2 edition, 2003

    work page 2003

  59. [59]

    Reppen, J

    A. Reppen, J. Rochet, and H. Soner. Optimal dividend pol icies with random profitability. Math. Finance, 30:228–259, 2020

    work page 2020

  60. [60]

    Revuz and M

    D. Revuz and M. Yor. Continuous martingales and Brownian motion , volume 293. Springer Science & Business Media, 2013

    work page 2013

  61. [61]

    F. Riedel. Optimal stopping with multiple priors. Econometrica, 77(3):857–908, 2009

    work page 2009

  62. [62]

    L. C. G. Rogers and D. Williams. Diffusions, Markov Processes and Martingales , volume 2 of Cambridge Mathematical Library . Cambridge University Press, 2 edition, 2000

    work page 2000

  63. [63]

    Schroder and C

    M. Schroder and C. Skiadas. Optimal consumption and por tfolio selection with stochastic dif- ferential utility. J. Econ. Theory , 89(1):68–126, 1999

    work page 1999

  64. [64]

    S. E. Shreve, J. P. Lehoczky, and D. P. Gaver. Optimal con sumption for general diffusions with absorbing and reflecting barriers. SIAM J. Control Optim. , 22(1):55–75, 1984

    work page 1984

  65. [65]

    S. E. Shreve and H. M. Soner. A free boundary problem rela ted to singular stochastic control. Applied stochastic analysis (London, 1989) , 5:265–301, 1991

    work page 1989

  66. [66]

    C. Skiadas. Robust control and recursive utility. Finance Stoch., 7:475–489, 2003

    work page 2003

  67. [67]

    H. M. Soner, S. E. Shreve, and N. El Karoui. A free boundar y problem related to singular stochastic control: the parabolic case. Commun. Partial Differ. Equ. , 16(2-3):373–424, 1991

    work page 1991

  68. [68]

    H. Xing. Consumption–investment optimization with Ep stein–Zin utility in incomplete markets. Finance Stoch., 21:227–262, 2017

    work page 2017

  69. [69]

    B. Yi, Z. Li, F. G. Viens, and Y. Zeng. Robust optimal cont rol for an insurer with reinsurance and investment under heston’s stochastic volatility model . Insur. Math. Econ. , 53(3):601–614, 2013

    work page 2013

  70. [70]

    Yin and Y

    C. Yin and Y. W en. Optimal dividend problem with a termin al value for spectrally positive Lévy processes. Insur. Math. Econ. , 53(3):769–773, 2013. 44 KEXIN CHEN, KYUNGHYUN PARK, AND HOI YING WONG Department of Applied Mathematics, The Hong Kong Polytechn ic University, Hong Kong Email address : kexinchen@polyu.edu.hk Division of Mathematical Sciences, ...

    work page 2013