Pith. sign in

REVIEW 2 major objections 4 minor 3 cited by

Mean Field Control with Poissonian Common Noise: A Pathwise Compactification Approach

T0 review · 2 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper proves that, under standard Lipschitz and growth conditions, the mean-field control problem with Poissonian common noise always has an optimal relaxed control, constructed by freezing each sample path of the noise, solving the…

desk verdict Novel pathwise compactification for Poissonian common noise, but the aggregation step in Theorem 4.1(iii) has a genuine non-anticipation gap that blocks the main existence theorem. read the letter →

arxiv 2505.23441 v4 pith:AVZMTQUA submitted 2025-05-29 math.OC math.PR

Lijun Bo , Jingfei Wang , Xiaoli Wei , Xiang Yu This is my paper
classification math.OCmath.PR MSC 49N8060H3093E2060G07
keywords meanfieldcontrolgamePoissoniancommonnoisepathwiseformulationcompactificationapproachrelaxedstrongequilibrium
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper claims that mean-field control problems with a Poissonian common noise always admit an optimal relaxed control, provided the noise has finite intensity and the coefficients satisfy standard Lipschitz and growth conditions. The proof works by freezing one sample path of the common noise at a time, solving a deterministic-jump control problem with the classical compactification method, and then stitching the pathwise solutions together over all sample paths. The payoff of the argument is that the resulting optimal control is adapted to the common noise, and the same construction yields a strong mean-field equilibrium for the corresponding mean-field game. If the claim is right, Poisson shocks such as crises or policy jumps do not break the existence theory for cooperative or competitive mean-field models.

What carries the argument

The load-bearing object is the pathwise formulation: for each fixed sample path $\omega_1$ of the Poisson noise, define admissible relaxed controls $P^{\omega_1}\in R(\omega_1)$ through a martingale problem in which the noise term becomes a finite sum of deterministic jumps. The argument then uses two devices: a compactification in the Skorokhod topology on the pathwise problems to obtain a measurable selection of optimal controls, and a pathwise superposition principle that constructs, from any measure flow satisfying the Fokker–Planck equation with deterministic jumps, a martingale-measure representation with the same marginals. Concatenation over the deterministic jump times is the technical glue that makes the superposition principle work, and the identity $\inf_{\bar P\in R} J(\bar P)=\int_{\Omega_1}\inf_{P^{\omega_1}\in R(\omega_1)} J(\omega_1,P^{\omega_1}) P^1(d\omega_1)$ links the pathwise problems to the original one.

What would settle it

Build a model satisfying Assumption 1 and find a pathwise measure-valued control $(\mu^{\omega_1},\hat\alpha^{\omega_1})\in R_{FP}(\omega_1)$ with deterministic jumps for which no pathwise relaxed control $P^{\omega_1}\in R(\omega_1)$ has marginals $\mu^{\omega_1}$ and relaxed action $\hat\alpha^{\omega_1}$; this would refute the superposition principle in Theorem 4.1(ii) and break the aggregation step. A cheaper check is to add a terminal cost to the objective and exhibit an example where the value with terminal cost is strictly below the pathwise-aggregated value, since the compactification step explicitly drops terminal costs.

Watch

Extended reading notes

Core claim

The central discovery is that the lack of compactness and continuity caused by common noise can be bypassed for Poissonian common noise by using the point-process representation of the noise. Since the intensity is finite, every sample path has finitely many jumps, so freezing a path reduces the problem to a classical relaxed control problem with deterministic jump times. The paper proves a pathwise superposition principle: every measure-valued control satisfying the Fokker–Planck equation with deterministic jumps can be realized by a pathwise relaxed control. Aggregating the resulting optimal pathwise controls against the law of the common noise yields an admissible relaxed control $\bar P^*(d\omega,d\omega_1)=P^{\omega_1}_*(d\omega)P^1(d\omega_1)$ that attains the original value, proving Theorem 2.7. The same aggregation produces a strong mean-field equilibrium in the game version, Theorem 5.7.

Load-bearing premise

The argument collapses if the common noise is not a finite-intensity Poisson process, because freezing a path is only useful when each path has finitely many jumps, and the pathwise integrals and finite concatenations use that finiteness.

Editorial extensions

If this is right

  • Under Assumption 1, the optimal relaxed control set $R_{opt}(\lambda)$ is nonempty, so the MFC problem has a solution in the weak, relaxed sense.
  • If Assumption 2 (convexity) also holds, a strict control can realize the optimum, as stated in Corollary 2.8.
  • The value of the original problem splits as an integral over common-noise paths of pathwise values, giving a scenario-by-scenario certificate of optimality.
  • For the MFG extension, a strong mean-field equilibrium exists: the equilibrium measure flow is adapted to the filtration of the Poisson common noise.
  • Because the construction is pathwise, the optimal solution is decomposable into one optimal control per noise path, chosen measurably in the noise path.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural extension is to allow the jump coefficient $\gamma$ to depend on the control; the paper notes this changes the Fokker–Planck equation and leaves it open, so a testable next step is whether the superposition principle survives with randomized jumps at the deterministic times.
  • The finite-intensity restriction suggests an approximation route for infinite-activity common noise: approximate the Poisson random measure by finite-intensity compound Poisson measures and check whether the pathwise value functions converge to a limit that solves the original problem.
  • The scenario-decomposition viewpoint may support numerical schemes: solve deterministic-jump pathwise problems per sampled noise path in parallel and average, which would give a practical algorithm if the measurable selection can be computed approximately.
  • Because the method yields strong, common-noise-adapted equilibria without the discretization step used for Brownian common noise, it may be worth testing whether a similar freezing argument works for other càdlàg common noises whose paths have finite activity, such as compound renewal processes.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 4 minor

Summary. The paper develops a pathwise compactification method for mean-field control (MFC) and mean-field games (MFG) under Poissonian common noise with finite intensity. The authors freeze a sample path of the common noise, solve an auxiliary relaxed control problem with deterministic jumps, and then aggregate the resulting pathwise optimal controls over the common-noise distribution. The main claims are the existence of an optimal relaxed control for the MFC problem (Theorem 2.7) and the existence of a strong mean-field equilibrium for the MFG extension (Theorem 5.7), both resting on the equivalence (29) in Theorem 4.1(iii). The paper also establishes a pathwise superposition principle (Theorem 4.1(ii)) via concatenation techniques.

Significance. If the main theorem is correct, the paper offers a new technique that avoids the discretization of common noise and produces common-noise-adapted (strong) mean-field equilibria, a significant step beyond the weak equilibria obtained by prior compactification approaches. The pathwise superposition principle for deterministic jumps is also novel and of independent interest. However, the central aggregation step in Theorem 4.1(iii) has a load-bearing gap concerning causality of the pathwise optimizers with respect to the common-noise filtration; the stated results are therefore not fully established. The paper provides detailed compactness and continuity arguments and clearly identifies the role of finite jump intensity, which are valuable features even if the main equivalence needs repair.

major comments (2)
  1. [Theorem 4.1(iii), Eq. (35)-(36)] The proof that the aggregated measure P_bar_* belongs to R assumes that the measurable selection omega1 -> P^{omega1}_* produces a flow mu^{omega1}_t which is F^1_t-measurable for each t. This non-anticipation property is not proved and is in fact not implied by the pathwise formulation: Definition 3.1 fixes the entire path omega1 and imposes no causality constraint on how P^{omega1}_* may depend on future jumps. In the verification of Definition 2.4(iii), the set B2 involves increments omega1((t,s] x A) with s>t, which are not in F^1_t. The displayed equality P_1(B2) P_bar_*(X_t in C, N in B1) = P_bar_*(X_t in C, N in B) requires that X_t be independent of future common-noise increments under P_bar_*, which is exactly the unproved non-anticipation property. In general, L^{P_bar_*}(X_t|F^1_t) is the conditional expectation of the kernel P^{omega1}_*(X_t in .) given F^1_t, and it equals mu^{omega1}_t only if the kernel is F^1_t-measurable. Consequently, the constructed P_bar_* need not satisfy Definition 2.4(iii), and the key equivalence (29) is not established. This gap directly affects Theorem 2.7.
  2. [Theorem 5.7] The extension to mean-field games inherits the same causality gap. The proof asserts that the consistency condition bar_mu^*_t = L^{P_bar_*}(X_t|F^1_t) follows from the pathwise consistency condition together with the compatibility condition (2), but this again requires that the measurable family of pathwise MFEs (mu^{omega1}, P^{omega1}_*) be non-anticipative in omega1. The argument in the sketch does not establish that the selected pathwise equilibria depend on omega1 only through F^1_t. Without this, the aggregated pair (bar_mu^*, P_bar_*) need not be a strong MFE in the sense of Definition 5.3.
minor comments (4)
  1. [Eq. (36)] In the first line of the display, the expression 'M^{omega1,P^{omega1}_*}_phi(t) - M^{omega1,P^{omega1}_*}_phi(t)' appears twice with the same argument t; it should be the difference between the values at t and s.
  2. [General] There are several typographical slips: 'Portmaneau' should be 'Portmanteau', 'Arzela-Ascoli' should be 'Arzelà–Ascoli', and 'may result in an challenge' should be 'may result in a challenge'.
  3. [Lemma 3.10] In the proof, the phrase 'in the 3rd inequality' should likely be 'in the 3rd equality', since the text refers to an application of the Portmanteau theorem.
  4. [Section 6.2] The notation P1 ⊗t2 P · is used without an explicit definition of the domain of the kernel; a short clarification would help readability.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the central aggregation equivalence is argued directly, not by assuming the target result; the only self-citation appears in the narrative and is not load-bearing in the proof.

full rationale

The paper's derivation chain is not circular. Step 1 establishes existence of pathwise optimal relaxed controls by standard compactification in the Skorokhod topology, using external results (Haussmann-Suo, Trevisan) and containing no fitted parameters. Step 2 does not assume the key equivalence (29); it proves the two inequalities separately. The inequality 'inf over original relaxed controls >= pathwise aggregate' is derived in the proof of Theorem 4.1(iii) through the Fokker-Planck disintegration (34), and the reverse inequality is attempted by explicitly constructing P_bar*(dω,dω1)=P^ω1_*(dω)P1(dω1) and verifying the admissibility conditions of Definition 2.4. The one self-citation in the introduction, 'Lemma 6.14 in [7] implies that the value function of the original problem with common noise is less than that of the pathwise formulation,' is not actually used in the proof of Theorem 4.1(iii), which supplies both bounds directly; hence this self-citation is not load-bearing. The stated limitations (dropping terminal cost in Remark 2.1, uncontrolled jump coefficient in Remark 3.8, finite intensity in Remark 4.3) narrow the theorem but do not import the conclusion. The aggregation step does contain a possible non-causality gap: the proof factors P1(B2) for a B2 involving common-noise increments after time t out of a conditional expectation given F^1_t, which would require the pathwise optimizers to be non-anticipative in the common noise; however, this is a correctness concern about an unproved independence, not a circular reduction where an equation is defined in terms of the target result or a fitted input is renamed a prediction. Accordingly, the score reflects the minor non-load-bearing self-citation rather than any circularity of the main derivation.

Assumptions & free parameters 0 free parameters · 8 assumptions · 0 invented entities

The proofs rest on standard martingale-problem and superposition tools plus the explicit modelling assumptions: finite Poisson intensity, running cost only, and uncontrolled jump coefficient. No free parameters are fitted, and no new physical or mathematical entities are postulated.

assumptions (8)
  • domain assumption Finite intensity of the common-noise Poisson random measure, so each sample path has finitely many jumps over [0,T].
    Used throughout; enables pathwise stochastic integrals and finite concatenation. Stated in Section 2 and Remark 4.3.
  • domain assumption Coefficients b, sigma, f, gamma satisfy continuity, Lipschitz, growth, and integrability conditions (Assumption 1).
    Used for compactness of admissible sets, cost continuity, and moment estimates; Section 2.
  • domain assumption The jump coefficient gamma is uncontrolled, i.e. it does not depend on the control variable u.
    Remark 3.8 states that controlled jumps break the Fokker-Planck and superposition equivalence; this is a real restriction of the theorems.
  • domain assumption The objective contains no terminal cost.
    Remark 2.1 removes terminal costs because Skorokhod convergence does not imply convergence at T; the compactification relies on this.
  • domain assumption The control space U is compact.
    Makes the relaxed-control space Q compact and Polish; stated in Section 2.
  • standard math Classical superposition principle for linear Fokker-Planck equations and Stroock-Varadhan martingale problem and selection theorems.
    External theorems cited in Theorem 4.1(ii) and Proposition 3.9, in particular Trevisan [42] and Stroock-Varadhan [39].
  • standard math Compatibility condition for the Poisson filtration, EP[1_D | F_N_t] = EP[1_D | F_N_T] for all relevant D.
    Used to identify conditional laws in Lemma 2.5 and Section 5; a standard property of finite-intensity Poisson random measures.
  • standard math Weak* convergence of finite counting measures implies eventual equality of the corresponding point functions.
    Used in Remark 2.3 and Lemma 3.12 to prove closedness of the graph of omega1 to R_opt(omega1).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mean Field Control with Poissonian Common Noise: A Pathwise Compactification Approach." pith.science (2026). https://pith.science/paper/AVZMTQUA

@misc{pith2026250523441,
  author       = {Pith},
  title        = {Pith review of: Mean Field Control with Poissonian Common Noise: A Pathwise Compactification Approach},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AVZMTQUA}},
  note         = {Machine review of arXiv:2505.23441}
}
read the original abstract

This paper contributes to the compactification approach to study mean-field control problems with Poissonian common noise. To overcome the lack of compactness and continuity issues caused by common noise, we exploit the point process representation of the Poisson random measure with finite intensity and propose a pathwise formulation in a two-step procedure by freezing a sample path of the common noise. In the first step, we establish the existence of optimal relaxed controls in the pathwise formulation as if common noise is absent, but with finite deterministic jumping times. The second step plays the key role in our approach, which is to aggregate the optimal solutions in the pathwise formulation over all sample paths of common noise and show that it yields an optimal solution in the original model. To this end, with the help of concatenation techniques, we first develop a pathwise superposition principle in the model with deterministic jumping times, drawing a relationship between the pathwise relaxed control problem and the pathwise measure-valued control problem. We then further bridge the equivalence among different problem formulations and verify that the constructed solution under aggregation is indeed optimal in the original problem.

Discussion (0). Sign in to comment.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Mean-field games with rough common noise: the linear-quadratic case

    math.PR 2026-02 unverdicted novelty 7.0 of 10

    The paper establishes existence, uniqueness, and stability for linear-quadratic mean-field games with rough common noise via a novel Volterra formulation and rough FBSDEs.

  2. Mean-Variance Stackelberg Games with Asymmetric Information

    q-fin.MF 2025-09 conditional novelty 6.0 of 10

    In a two-investor mean-variance Stackelberg game with asymmetric information, the leader's equilibrium randomized strategy is Gaussian and the follower's strategy depends linearly on the leader's observed trades.

  3. Ergodicity of conditional McKean-Vlasov jump diffusions

    math.PR 2025-09 conditional novelty 6.0 of 10

    Conditional McKean-Vlasov jump diffusions are exponentially contractive in law, and the contraction rate improves as jump noise intensity grows.

Reference graph

Works this paper leans on

43 extracted references · 37 canonical work pages · cited by 3 Pith papers

  1. [1]

    (2016): Wellposedness of mean field games with common noise under a weak monotonicity condition.SIAM J

    Ahuja, S. (2016): Wellposedness of mean field games with common noise under a weak monotonicity condition.SIAM J. Contr. Optim.54, 30-48

    work page 2016

  2. [2]

    and Border, K.C

    Aliprantis, C.D. and Border, K.C. (2006):Infinite Dimensional Analysis: A Hitchhiker’s Guide, 3rd Edition.Springer-Verlag, Berlin. 26

    work page 2006

  3. [3]

    and Touzi, N

    Barrasso, A. and Touzi, N. (2022): Controlled diffusion mean field games with common noise and McKean-Vlasov second order backward SDEs.Theor. Probab. Appl.66(4), 613-639

    work page 2022

  4. [4]

    and Persio, L.D

    Benazzoli, C., Campi, L. and Persio, L.D. (2020): Mean field games with controlled jump–diffusion dynamics: Existence results and an illiquid interbank market.Stoch. Process. Appl.130, 6927-6964

    work page 2020

  5. [5]

    and Yu, X

    Bo, L., Wang, J., Wei, X. and Yu, X. (2024): Extended mean-field control problems with Poissonian common noise: Stochastic maximum principle and Hamiltonian-Jacobi-Bellman equation. Preprint, available athttps://arxiv.org/abs/2407.05356

    arXiv 2024

  6. [6]

    Constrained mean-field control with singular controls: Existence, stochastic maximum principle and constrained FBSDE

    Bo, L., Wang, J. and Yu, X. (2025a): Constrained mean-field control with singular control: Existence, stochastic maximum principle and constrained FBSDE. Preprint, available athttps://arxiv.org/ abs/2501.12731

    work page Pith review arXiv 2025

  7. [7]

    and Yu, X

    Bo, L., Wang, J. and Yu, X. (2025b): Mean field game of controls with state reflections: Existence and limit Theory. Preprint, available athttps://arxiv.org/abs/2503.03253

    arXiv 2025

  8. [8]

    and Campi, L

    Burzoni, M. and Campi, L. (2023): Mean field games with absorption and common noise with a model of bank run.Stoch. Process. Appl.164, 206-241

    work page 2023

Show all 43 references
  1. [9]

    and Fischer, M

    Campi, L. and Fischer, M. (2018): N-player games and mean-field games with absorption.Ann. Appl. Probab.28, 2188–2242

    2018

  2. [10]

    and Lions, P.L

    Cardaliaguet, P., Delarue, F., Larsy, J.M. and Lions, P.L. (2019):The Master Equation and the Convergence Problem in Mean Field Games: (AMS-201). Princeton University Press, Princeton

    2019

  3. [11]

    and Delarue, F

    Carmona, R. and Delarue, F. (2017):Probabilistic Theory of Mean Field Games with Applications. Volume II: Mean Field Games with Common Noise and Master Equations, Springer-Verlag, New York

    2017

  4. [12]

    and Lacker, D

    Carmona, R., Delarue, F. and Lacker, D. (2016): Mean field games with common noise.Ann. Probab.44, 3740-3803

    2016

  5. [13]

    and Fischer, M

    Cecchin, A. and Fischer, M. (2020): Probabilistic approach to finite state mean field games.Appl. Math. Optim.81, 253-300

    2020

  6. [14]

    and Pham, H

    Denkert, R., Kharroubi, I. and Pham, H. (2024): A randomisation method for mean-field control problems with common noise. Preprint, available athttps://arXiv:2412.20782

    2024 arXiv

  7. [15]

    and Tan, X

    Djete, M.F., Possamai, D. and Tan, X. (2022a): McKean-Vlasov optimal control: Limit theory and equivalence between different formulations.Math. Oper. Res.47, 2891-2930

    2022

  8. [16]

    and Tan, X

    Djete, M.F., Possama ¨ ı, D. and Tan, X. (2022b): McKean-Vlasov optimal control: the dynamic programming principle.Ann. Appl. Probab.50(2), 791-833

    2022

  9. [17]

    and Garleanu, N

    Duffie, D. and Garleanu, N. (2001): Risk and valuation of collateralized debt obligations.Financ. Anal. J.57(1), 41-59

    2001

  10. [18]

    and Sun, L.H

    Fouque, J.P., Carmona, R. and Sun, L.H. (2015): Mean field games and systemic risk.Commun. Math. Sci.13(4), 911-933

    2015

  11. [19]

    and Horst, U

    Fu, G. and Horst, U. (2017): Mean field games with singular controls.SIAM J. Contr. Optim.55, 3833-3868

    2017

  12. [20]

    and Suo, W.L

    Haussmann, U.G. and Suo, W.L. (1995): Singular optimal stochastic controls I: Existence.SIAM J. Contr. Optim.33, 916-936

    1995

  13. [21]

    and Lepeltier, J.P

    Haussmann, U.G. and Lepeltier, J.P. (1990): On the existence of optimal controls.SIAM J. Contr. Optim.28, 851-902

    1990

  14. [22]

    and Ricalde-Guerrero, J.H

    Hern´ andez-Hern´ andez, D. and Ricalde-Guerrero, J.H. (2023): Conditional McKean-Vlasov dif- ferential equations with common Poissonian noise: Propagation of chaos. Preprint, available at https://arxiv.org/abs/2308.11564 27

    2023 arXiv

  15. [23]

    (2024): Mean-field games with common Poissonian noise: A maximum principle approach

    Hern´ andez-Hern´ andez, D., and Ricalde-Guerrero, J.H. (2024): Mean-field games with common Poissonian noise: A maximum principle approach. Preprint, available athttps://arXiv:2401.10952

    2024 arXiv

  16. [24]

    and Caines, P.E

    Huang, M., Malham´ e, R.P. and Caines, P.E. (2006): Large population stochastic dynamic games closed-loop McKean-Vlasov systems and the nash certainty equivalence principle.Commun. Inform. Syst.6(3), 221-252

    2006

  17. [25]

    and Watanabe, S

    Ikeda, N. and Watanabe, S. (1989):Stochastic Differential Equations and Diffusion Processes, 2nd Edition.North-Holland Mathematical Library, UK

    1989

  18. [26]

    and M´ emin, J

    Jacod, J. and M´ emin, J. (1981):Sur un type de convergence interm´ ediaire enter la convergence en loi et la convergence en probabilit´ e.in S´ enibaire de Probabilit´ e XV, Lecture Notes in Math. 850, Springer-Verlag, 529-540

    1981

  19. [27]

    I., Nguyen, N.D.H

    Karoui, E. I., Nguyen, N.D.H. and Jeanblanc-Picqu´ e, M. (1987): Compactification methods in the control of degenerate diffusions: existence of an optimal control.Stochastics.20, 169-219

    1987

  20. [28]

    (2015): Mean field games via controlled martingale problems: Existence of Markovian equilibria.Stoch

    Lacker, D. (2015): Mean field games via controlled martingale problems: Existence of Markovian equilibria.Stoch. Process. Appl.125(7), 2856-2894

    2015

  21. [29]

    (2017): Limit theory for controlled McKean-Vlasov dynamics.SIAM J

    Lacker, D. (2017): Limit theory for controlled McKean-Vlasov dynamics.SIAM J. Contr. Optim. 55, 1641-1642

    2017

  22. [30]

    and Zhang, J

    Lacker, D., Shkolnikov, M. and Zhang, J. (2023): Superposition and mimicking theorems for condi- tional McKean-Vlasov equations.J. Eur. Math. Soc.25, 3229-3288

    2023

  23. [31]

    (2018): Mean field game and interacting particle systems

    Lacker, D. (2018): Mean field game and interacting particle systems. Lecture Notes available at https://www.columbia.edu/~dl3133/MFGSpring2018.pdf

    2018

  24. [32]

    and Lions, P.L

    Lasry, J.M. and Lions, P.L. (2007): Mean field games.Japanese J. Math.2(1), 229-260

    2007

  25. [33]

    (1978): Measurable selections and the uniformization of souslin sets.Amer

    Leese, S.J. (1978): Measurable selections and the uniformization of souslin sets.Amer. J. Math. 100, 19-41

    1978

  26. [34]

    and Zhang, K

    Liang, Z., Yu, X. and Zhang, K. (2024): On time-inconsistent extended mean-field control problems with common noise. Preprint, available at:https://arxiv.org/abs/2409.07219

    2024 arXiv

  27. [35]

    and McNeil, A.J

    Lindskog, F. and McNeil, A.J. (2003): Common Poisson shock models: applications to insurance and credit risk modelling.ASTIN Bulletin: J. IAA.33(2), 209-38

    2003

  28. [36]

    and Zhang, J

    Mou, C. and Zhang, J. (2024): Mean field game master equations with anti-monotonicity conditions. J. Eur. Math. Soc.11, 4469-4499

    2024

  29. [37]

    and Wei, X

    Pham, H. and Wei, X. (2017): Dynamic programming for optimal control of stochastic McKean- Vlasov dynamics.SIAM J. Contr. Optim.55(2), 1069–1101

    2017

  30. [38]

    and Zhang, X

    Rockner, M., Xie, L. and Zhang, X. (2020): Superposition principle for nonlocal Fokker-Planck- Kolmogorov operators.Probab. Theor. Relat. Field.178, 699-733

    2020

  31. [39]

    and Varadhan, S.R.S

    Stroock, W.D. and Varadhan, S.R.S. (1979):Multidimensional Diffusion Processes. Springer-Verlag, New York

    1979

  32. [40]

    and Wang, S

    Tangpi, L. and Wang, S. (2024): Optimal bubble riding with price-dependent entry: a mean field game of controls with common noise.Math. Finan. Econ.18, 275-312

    2024

  33. [41]

    and Wang, S

    Tangpi, L. and Wang, S. (2025): Mean field games with common noise via Malliavin calculus. Electron. Commun. Probab.30, 1-14

    2025

  34. [42]

    (2016): Well-posedness of multidimensional diffusion process with differentiable coef- ficients.Eletron

    Trevisan, D. (2016): Well-posedness of multidimensional diffusion process with differentiable coef- ficients.Eletron. J. Probab.22, 1-41

    2016

  35. [43]

    and Zhang, J

    Zhou, J., Touzi, N. and Zhang, J. (2024): Viscosity solutions for HJB equations on the process space: Application to mean field control with common noise. Preprint, available at arXiv:2401.04920. 28

    2024 arXiv

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.