Mean-field model for pollution abatement via cap and trade mechanism

Giacomo Lanaro; Ofelia Bonesini

REVIEW 4 major objections 5 minor 27 references

Mean-field model for pollution abatement via cap and trade mechanism

T0 review · 4 major / 5 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read This paper derives an optimal permit-issuance rule for a mean-field cap-and-trade economy, and a calibrated scenario shows it cuts cumulative emissions to 50% of the no-policy level while losing about 9% of aggregate capital.

desk verdict A useful mean-field extension making the regulator an optimizer in auction-based cap-and-trade, but the separation lemma and control constraint need theorem-level fixes. read the letter →

arxiv 2607.22638 v1 pith:WFFUFU26 submitted 2026-06-22 math.OC math.PR

Ofelia Bonesini , Giacomo Lanaro This is my paper · ORCID

classification math.OCmath.PR MSC 49N1060H3091B7093E20

keywords mean-fieldgamescapandtradeemissiontradingsystemsoptimalcontrolFBSDERiccatiequationsAKproductionmodelpermitauction

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

The paper tries to establish that a cap-and-trade regulator can be treated as an active optimizer who auctions new emission permits over time, rather than a passive rule-setter. In the mean-field limit of a large economy of AK-type firms, the paper reduces the regulator's problem to a linear-quadratic stochastic control problem whose solution is characterized by a coupled FBSDE and Riccati equations. A calibrated numerical scenario shows the resulting policy meeting a 50% cumulative-emission-reduction target while holding the loss in aggregate capital near 9%, and flipping the energy mix so green investment exceeds fossil investment after roughly one year. The point of the exercise is that the emissions-output trade-off in an ETS can be quantified and optimized through a dynamic auction supply policy.

What carries the argument

The load-bearing object is the mean-field market-clearing price ϖ = −Ȳ + 2c_{2,x}β, which transmits the regulator's permit issuance into the firms' FBSDE equilibrium. The separation Lemma 4.1, which shows the filtering error is independent of β, collapses the regulator's problem onto the conditional-mean dynamics (4.2), and the Riccati ODE system then yields the optimal β in feedback form via (4.16).

What would settle it

Solve the coupled FBSDE (3.7) numerically for two different admissible β paths sharing the same Brownian paths and compare the distributions of (X−X̄, Y−Ȳ); if they differ, Lemma 4.1 fails and the reduction to (4.2) is invalid. A cheaper check: compute the unconstrained optimal β from (4.16) and see whether it ever becomes positive, which would violate the regulator's no-withdrawal constraint in Remark 2.2.

Watch

Extended reading notes

Core claim

The central claim is that, once the firms' Nash equilibrium is passed to the mean-field limit, the regulator's optimal auction-supply policy β can be computed from the conditional means of the firm state and adjoint process. Lemma 4.1 isolates the filtering error (X−X̄, Y−Ȳ) from the control, so the regulator's objective separates into a term driven by the conditional means plus a β-independent error term. This leaves a Markovian linear-quadratic control problem in (X̄, B, Ψ) whose solution is a feedback rule assembled from Riccati ODEs. The paper then simulates a single scenario and reports that the optimal β is strictly negative (the regulator is a net supplier), keeps market clearing to w

Load-bearing premise

The argument relies on the unproved assumption that the firms' equilibrium system has a unique solution, so that the part of the state the regulator cannot see is unaffected by how many permits she issues.

Editorial extensions

If this is right

Regulators can, in principle, compute a time-dependent permit-auction supply that meets a preset emissions target while limiting aggregate output loss, using only common-noise information.
The mean-field limit makes the equilibrium price adapted to the common shock, removing the intractable dependence on every firm's idiosyncratic noise.
The optimal policy can be implemented by solving a set of Riccati ODEs and one SDE, with the permit price and firm controls given in closed feedback form.
In the calibrated scenario, the optimal policy satisfies market clearing to numerical precision and flips the energy mix: green investment overtakes fossil investment after about one year.

Reading between the lines

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

One implication the authors leave implicit: the same Riccati machinery would extend to multi-population economies (their Remark 3.1), so sector-differentiated ETS design could be tackled with the same separation argument.
A testable extension would be to sweep the target θ and the penalty weight a(t) to trace the emissions-output Pareto frontier; the paper's 50%/9% point is a single scenario, not a frontier.
The numerical claim rests on a single common-noise path; averaging over many W⁰ realizations would show whether the 9% capital-loss figure is typical or path-dependent.
The mismatch between the stated constraint β≤0 and the unconstrained first-order condition means the optimal policy may ask the regulator to withdraw permits in some states; projecting onto admissible controls would alter the reported outcomes.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Two things before you dive in. First, the new contribution is real: the paper closes the loop on DSLL24 by making the regulator an active optimizer of auction-based permit supply in the mean-field limit, with the price emerging endogenously. Second, the central separation argument (Lemma 4.1) is conditional on strong well-posedness of the coupled FBSDE, which is not proved, and the control constraint on β is stated inconsistently. Neither gap looks fatal, but both need theorem-level support before the characterization is established.

What is actually new: the regulator's problem is solved from stated assumptions using standard LQ/FBSDE tools, with no reliance on the authors' own prior results. The rewriting of the objective in Section 2.3 is careful, and the reduction to a finite-dimensional Riccati system is clean when the hypotheses hold. The numerical illustration, while single-scenario, shows a plausible trade-off: the optimal policy hits the 50% reduction target while limiting capital loss to about 9%, and green investment overtakes fossil investment near t≈1. The paper is honest, explicitly flagging where existence and uniqueness are assumed.

The stress-test note lands. Lemma 4.1's proof is conditional: 'If uniqueness of strong solutions holds, by Yamada-Watanabe...' That is a big if. The error FBSDE and the Riccati ODEs need well-posedness conditions. If the error dynamics has no strong solution, the separation into (4.1)–(4.2) collapses, and the FOC (4.16) would not represent the original problem. Also, Remark 2.2 requires β≤0 (regulator can only issue permits), but (3.9) optimizes over [0,∞), and the FOC is unconstrained. The numerical β-hat is negative, so the intended domain is probably (-∞,0], but the inconsistency is concrete and must be resolved. On the numerics: θ is defined as 50% of no-policy emissions, so the 50% reduction is an input target, not an independent prediction. That's fine for an illustration but should be labeled as such.

Who is this for: anyone working on mean-field games for environmental regulation or auction-based ETS design. The model is a plausible extension of DSLL24, and the gaps are addressable. It deserves a serious referee, not a desk reject.

My recommendation: send it to peer review with a request for major revision, asking for well-posedness results, a consistent control constraint, and a more careful numerical section.

Referee Report

4 major / 5 minor

Summary. The paper studies a mean-field model of competitive firms with AK production technology and emissions, interacting through a permit market with a regulator who controls permit supply via a dynamic auction. The authors formulate the regulator as a mean-field control problem whose state is the FBSDE describing firms' equilibrium, reduce it via a filtering argument to a lower-dimensional LQ problem, and derive candidate optimal policies through Riccati ODEs. Numerical experiments claim that the optimal policy achieves a 50% reduction in cumulative emissions relative to a no-policy benchmark with a capital loss of about 9%, and induces green investment to overtake fossil investment around t≈1.

Significance. If the theoretical characterization is made rigorous, the paper would provide a genuinely useful and tractable framework for optimal permit-supply design in large emission trading systems, going beyond the existing literature where the regulator is usually passive. The authors build on established mean-field game tools (Yamada–Watanabe representation, LQ-FBSDE reduction) and provide a concrete numerical implementation route in Appendix B. The main value is the closed-form reduction of a complex Stackelberg-type mean-field problem to Riccati equations, which is a significant methodological contribution. However, the central separation lemma and the constraint handling are currently conditional and inconsistent, so the contribution is not yet fully established.

major comments (4)

[§4, Lemma 4.1] The separation result that the filtering error (X−X̄, Y−Ȳ) is independent of β is the linchpin of the paper, but its proof is explicitly conditional: 'If uniqueness of strong solutions holds, by Yamada-Watanabe...'. Strong well-posedness of the coupled mean-field FBSDE (3.7) and of the error system is never proved. The manuscript itself concedes this in §3.2 ('We should prove the existence and uniqueness of the solution of (3.7) for every given process β') and §4 ('For the moment let us assume that the solutions of the ODEs and the SDE for Ψ exist uniquely'). Without these well-posedness results, the decomposition J0 = J̄0 + independent term and the ensuing FOC (4.16) are unsupported. Please provide either a theorem with explicit linear-quadratic conditions guaranteeing strong existence and uniqueness for (3.7) and the error system, or state the reduction as a formal derivation pending t
[§2.2.1, Remark 2.2 and §3.3, Eq. (3.9)] The admissible control set for β is stated inconsistently. Remark 2.2 requires β_t ≤ 0 (A = [−a,0] or (−∞,0]), so the regulator can only withdraw permits, while (3.9) optimizes over β ∈ H²(F0; [0,∞)). The first-order condition (4.16) is derived from an unconstrained minimization. Since the numerical solution reports β strictly negative, the intended constraint is likely β ≤ 0, but then the FOC must be replaced by the appropriate projection (or a Karush–Kuhn–Tucker condition). This is not merely cosmetic: with a one-sided constraint, the unconstrained candidate β̂ may not be optimal, and the value function and numerical outcomes would change if the constraint binds. Please reconcile the sign convention and state the true admissible set, then derive the optimality conditions accordingly.
[§5, Table 2 and surrounding text] The headline numerical claim—'the optimal policy achieves the desired 50% emission reduction'—is tautological because θ is defined as 50% of the no-policy emissions ('θ should be defined... e.g. as 50% of the CO2 emitted under the same economic conditions, but with no policy'). The terminal penalty and the regulator's objective push E_T toward θ, so reporting E_T/θ ≈ 1 is expected from the construction, not a finding. The capital-loss figure of about 9% is based on a single stochastic scenario with one parameter set, and no sensitivity analysis or confidence intervals are provided. Please present the results as calibration checks rather than predictive claims, and add robustness experiments (e.g., varying θ, b(t), a(t), and the noise seed) to support the qualitative findings such as the green-investment crossover at t≈1.
[§4, Eqs. (4.4)–(4.17)] The ansatz procedure assumes that the adjoint variable Y can be written as P_t(X_t−X̄_t)+Q_t X̄_t+Ψ_t+φ_t and later that the coupled system admits the Riccati form Y_t = Q_t X_t + q_t. These are strong structural assumptions. Even if the ODEs have solutions, one must verify that the resulting (X,Y) actually solve the original FBSDE (3.7) and that the candidate control (4.16) is admissible (in particular, satisfies the constraint on β). Currently the paper states 'For the moment let us assume that the solutions of the ODEs and the SDE for Ψ exist uniquely' without providing conditions. For a linear FBSDE, existence of a solution to the Riccati ODE (4.9) is essentially equivalent to well-posedness; please provide explicit conditions on the coefficient matrices (e.g., positive definiteness, symmetry, sufficient conditions for global solvability) and verify the verification theorem.

minor comments (5)

[Notation and typos] Several typos and unclear notations: 'c`adl`ag' in §Notation; the definition of γ5 is missing a minus sign in the first display of (3.8) (it appears with a minus only later); in (4.2) the term 2(Q1+Q2+q1)X̄_s mixes a vector q1 with matrix multiplication—should be 2(Q1+Q2)X̄_s + 2q1; the caption of Figure 2 says 'multiplied by' without specifying the factor.
[§3.3, Eq. (3.8)] The tower-property justification is terse. Please spell out why E[β_t E[Y_t | F^0_t]] = E[β_t Y_t] under β ∈ H²(F0) and why the terminal term B_T² can be kept as is; this will help readers follow the rewriting.
[§5, Table 1] The choice c2,e = (2×0.211)^{-1} and c2,x = (2×285.713)^{-1} looks oddly precise; please add a brief calibration source or explain the values. Also, the no-policy emissions EWPT are not defined explicitly in Table 2—state where they come from (Appendix A?).
[§4, after Eq. (4.16)] The derivation of the closed-form FBSDE system (the 'new system') involves many matrix definitions (O, M, N, T, U, F) that are introduced only later. Moving these definitions before the system or including a table of notation would improve readability.
[Appendix B] Step 1 writes the Riccati equations in a different notation from §4 (e.g., AY vs A_Y). Please unify the notation so the implementation can be checked against the theory without re-deriving every coefficient.

Circularity Check

1 steps flagged · score 2.0 of 10

Central FBSDE/Riccati derivation is self-contained; only the numerical '50% emission reduction' is the input target θ by construction.

self definitional [Section 5 (Numerical Experiments), Table 2 and following text; θ defined in Section 2.2.2 and Section 5 parameter paragraph]
"The regulator targets a 50% reduction of cumulative emissions relative to the no-policy benchmark, i.e. θ = 0.5 E^WP_T ... Under the optimal regulatory policy, cumulative emissions at T = 5 amount to approximately 50% of those recorded in the no-policy economy, confirming that the regulator successfully steers the economy to the prescribed target θ."

The regulator's objective (2.14)/(3.8) contains a terminal penalty (average emissions − θ)^2, and θ is set equal to 0.5 times the no-policy emissions. Thus the reported ratio E_T/E^WP_T ≈ 0.500 is the minimizer tracking its own input target, not an independent consequence of the model. The capital-loss and energy-mix findings are not circular because they are not fixed by θ.

full rationale

No load-bearing circularity in the derivation. The regulator's problem is solved from stated linear-quadratic primitives; the separation in Lemma 4.1 removes a filtering-error term whose dynamics contain no β, and the Riccati/FBSDE characterization follows by standard ansatz and maximum-principle arguments. The paper cites external sources (CD18, FT22a, DSLL24, etc.), not its own prior work, for the key structural tools. The only by-construction claim is numerical: θ is defined as 50% of the no-policy emissions, so the reported 50% reduction restates the target; this does not affect the derivation of the optimal issuance rule. The unproved well-posedness of the FBSDE (3.7) and of the Riccati ODEs is a correctness gap, not circularity.

Assumptions & free parameters 7 free parameters · 7 assumptions · 1 invented entities

The model relies on standard MFG/FBSDE theory plus several domain assumptions; the main ad hoc inputs are the penalty weight b(t)>c2,x, the ansatz separation, and the strong-uniqueness/well-posedness assumptions that are asserted but not proved.

free parameters (7)

θ (emission target) = 0.5×E_WP_T (no-policy terminal emissions)
Chosen as 50% of the no-policy benchmark; the terminal penalty steers emissions to it, so the numerical '50% reduction' is by construction.
b(t) issuance penalty = 1 + e^{t/2}
Ad hoc increasing function chosen to satisfy b(t)>c2,x and encode progressive stringency; no empirical calibration.
a(t) weight = 1/40
Chosen constant weight in regulator objective; no justification beyond calibration.
λ terminal penalty = 9.5e-3
Chosen in Table 1; controls firms' incentive to hold permits at T.
c2,x and c2,e quadratic costs = c2,x=1/(2×285.713), c2,e=1/(2×0.211)
Taken/adapted from [AB23] and [DSLL24]; numerical outcomes depend on them.
ρ (common-noise correlation) = 0.92
Calibrated 'to reflect high correlation of sectoral emissions with common macroeconomic shocks' citing [AB23]; no data shown.
Remaining Table 1 parameters (K0,E0,A0,af,ag,ae,δ,σK,σE,a,b,γ,A,c1,p,c2,g) = values in Table 1
Chosen so the no-policy economy reaches a 'realistic steady state'; no empirical fit documented.

assumptions (7)

ad hoc to paper Mean-field FBSDE (3.7) has a unique strong solution for every admissible β (existence/uniqueness)
Assumed by analogy to [DSLL24, Prop 4.2] and [CD13, CD15]; not proved in the paper. Lemma 4.1 and the Riccati reduction depend on it.
ad hoc to paper Filtering-error FBSDE has strong uniqueness, so Yamada-Watanabe makes the error independent of β
Condition stated in Lemma 4.1 proof ('If uniqueness of strong solutions holds'); no verification.
ad hoc to paper Ansatz (4.4) Y_t = P_t(X_t−X̄_t)+Q_t X̄_t+Ψ_t+φ_t and final Riccati ansatz Y_t=Q_t X_t+q_t are valid
Standard for LQ problems, but requires solvability of Riccati ODEs; paper assumes existence.
ad hoc to paper Regulator's control set/convexity: b(t)>c2,x and β unconstrained in F0
Convexity is guaranteed by b>c2,x; but the sign constraint on β is inconsistent, and the FOC (4.16) solves the unconstrained problem.
domain assumption Market clearing in mean-field limit: −β_s = E[α̂x_s|F0_s] and price ϖ_mf = −Ȳ_t + 2c2,xβ_t
Idiosyncratic noise averages out; standard in [FT22a/FT22b] but not proved here.
standard math Background theory: existence/uniqueness of MFGs with common noise [CDL16], Yamada-Watanabe [CD18], SMP for convex LQ problems
Invoked throughout.
domain assumption Firms are price takers (Assumption 2.1) and regulator observes only F0 in the limit
Economic assumption.

invented entities (1)

Real-time auction-based permit supply mechanism (regulator's β_t)
purpose: Mediates regulator-firm interaction through endogenous permit price; replaces direct allowance allocation
Institutional modeling assumption; no empirical data or outside-model falsifiable handle. The paper models EU-ETS-like auctions abstractly.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Mean-field model for pollution abatement via cap and trade mechanism." pith.science (2026). https://pith.science/paper/WFFUFU26

@misc{pith2026260722638,
  author       = {Pith},
  title        = {Pith review of: Mean-field model for pollution abatement via cap and trade mechanism},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WFFUFU26}},
  note         = {Machine review of arXiv:2607.22638}
}

read the original abstract

We consider a mean-field model of competitive firms operating under an AK production technology, where output is proportional to capital and production generates emissions. In this setting, we introduce a regulator whose objective is the reduction of cumulative emissions, in the spirit of Emission Trading Systems (ETS), where firms must hold and trade permits to cover their emissions in a regulated market. The regulator acts as a central planner and controls the supply of permits, balancing emission reduction and aggregate output. Permits are allocated through a dynamic auction mechanism that adjusts supply in real time to achieve both market efficiency and the regulator's long-term goals. The regulator and the firms interact through the endogenous permit price, which is determined by the regulator's policy and affects firms' optimal strategies via the market clearing condition. The regulator's optimal policy is derived within a Mean-Field Control (MFC) framework. Exploiting the linear-quadratic structure and the presence of common noise, we characterise the equilibrium via a system of coupled FBSDEs and associated Riccati equations. Our results provide a tractable characterisation of optimal permit allocation policies in large economies and offer insights into the design of efficient emission trading mechanisms.

Figures

Figures reproduced from arXiv: 2607.22638 by the authors.

**Figure 1.** Left: cumulative emissions Et under the optimal policy (top row) and the no-policy economy (bottom row). Right: capital Kt and its mean-field average Kt under the two regimes. All trajectories share the same Brownian realisation. Emission reduction and capital preservation. The first key result concerns the effectiveness of the optimal policy in achieving the emission target while preserving the productive capacity… view at source ↗

**Figure 2.** Left: equilibrium permit price ϖt . Right Top: Comparison between optimal issuance rate βbt of the regulator and the ETS trading rate for the representative company multiplied by. Right Down: Relative error between βb and −αb x . Equilibrium permit price and market clearing [PITH_FULL_IMAGE:figures/full_fig_p025_2.png] view at source ↗

**Figure 3.** Optimal production controls αb f t (blue) and αb g t (orange) in the nopolicy economy (left) and under optimal regulation (right). Energy-mix switching under regulation. Perhaps the most economically informative finding concerns the firms’ production strategy [PITH_FULL_IMAGE:figures/full_fig_p025_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 2 linked inside Pith

[1]

2024 43rd Chinese Control Conference (CCC) , pages=

Direct approach of linear-quadratic Stackelberg mean field games of backward-forward stochastic systems , author=. 2024 43rd Chinese Control Conference (CCC) , pages=. 2024 , organization=

2024
[2]

SIAM journal on Control and Optimization , volume=

Linear-quadratic optimal control problems for mean-field stochastic differential equations , author=. SIAM journal on Control and Optimization , volume=. 2013 , publisher=

2013
[3]

2020 39th Chinese Control Conference (CCC) , pages=

A linear quadratic Stackelberg game of backward stochastic differential equations with partial information , author=. 2020 39th Chinese Control Conference (CCC) , pages=. 2020 , organization=

2020
[4]

Advances in Difference Equations , volume=

Backward-forward linear-quadratic mean-field Stackelberg games , author=. Advances in Difference Equations , volume=. 2021 , publisher=

2021
[5]

General theory , author=

Statistics of random processes: I. General theory , author=. 2013 , publisher=

2013
[6]

2018 , publisher=

An introduction to optimal control of FBSDE with incomplete information , author=. 2018 , publisher=

2018
[7]

arXiv preprint arXiv:2412.11433 , year=

Mean field games of major-minor agents with recursive functionals , author=. arXiv preprint arXiv:2412.11433 , year=

arXiv
[8]

Mathematics of Operations Research , volume=

Optimization under rational expectations: A framework of fully coupled forward-backward stochastic linear quadratic systems , author=. Mathematics of Operations Research , volume=. 2023 , publisher=

2023

Show all 27 references

[9]

Probabilistic Theory of Mean Field Games with Applications II , volume =

Carmona, Ren. Probabilistic Theory of Mean Field Games with Applications II , volume =
[10]

Optimal dynamic regulation of carbon emissions market , volume =

A. Optimal dynamic regulation of carbon emissions market , volume =. Mathematical Finance , number =
[11]

Equilibrium price formation with a major player and its mean field limit , volume =

Fujii, Masaaki and Takahashi, Akihiko , date-added =. Equilibrium price formation with a major player and its mean field limit , volume =. ESAIM: Control, Optimisation and Calculus of Variations , pages =
[12]

Electronic Communications in Probability , number =

Ren. Electronic Communications in Probability , number =. 2013 , doi =

2013
[13]

Forward--backward stochastic differential equations and controlled McKean--Vlasov dynamics , year =

Carmona, Ren. Forward--backward stochastic differential equations and controlled McKean--Vlasov dynamics , year =. The Annals of Probability , pages =
[14]

Fully coupled forward-backward stochastic differential equations and applications to optimal control , volume =

Peng, Shige and Wu, Zhen , date-added =. Fully coupled forward-backward stochastic differential equations and applications to optimal control , volume =. SIAM Journal on Control and Optimization , number =
[15]

A mean field game approach to equilibrium pricing with market clearing condition , volume =

Fujii, Masaaki and Takahashi, Akihiko , date-added =. A mean field game approach to equilibrium pricing with market clearing condition , volume =. SIAM Journal on Control and Optimization , number =
[16]

Random economies with many interacting agents , volume =

F. Random economies with many interacting agents , volume =. Journal of mathematical economics , number =
[17]

Mean field and n-agent games for optimal investment under relative performance criteria , volume =

Lacker, Daniel and Zariphopoulou, Thaleia , date-added =. Mean field and n-agent games for optimal investment under relative performance criteria , volume =. Mathematical Finance , number =
[18]

Linear quadratic mean field type control and mean field games with common noise, with application to production of an exhaustible resource , volume =

Graber, P Jameson , date-added =. Linear quadratic mean field type control and mean field games with common noise, with application to production of an exhaustible resource , volume =. Applied Mathematics & Optimization , pages =
[19]

arXiv preprint arXiv:2407.12754 , title =

Del Sarto, Gianmarco and Leocata, Marta and Livieri, Giulia , date-added =. arXiv preprint arXiv:2407.12754 , title =

arXiv
[20]

Numerical solution of differential Riccati equations arising in optimal control problems for parabolic partial differential equations , author=. Ph. D. thesis, Escuela Politecnica Nacional , year=
[21]

SIAM Journal on Control and Optimization , volume=

Optimal stochastic control and carbon price formation , author=. SIAM Journal on Control and Optimization , volume=. 2009 , publisher=

2009
[22]

Siam Review , volume=

Market design for emission trading schemes , author=. Siam Review , volume=. 2010 , publisher=

2010
[23]

European Economic Review , volume=

Dynamic supply adjustment and banking under uncertainty in an emission trading scheme: the market stability reserve , author=. European Economic Review , volume=. 2019 , publisher=

2019
[24]

SIAM Review , volume=

Fracking, renewables, and mean field games , author=. SIAM Review , volume=. 2017 , publisher=

2017
[25]

Dynamic Games and Applications , volume=

Mean field models to regulate carbon emissions in electricity production , author=. Dynamic Games and Applications , volume=. 2022 , publisher=

2022
[26]

Mathematics and Financial Economics , volume=

Energy transition under scenario uncertainty: a mean-field game of stopping with common noise , author=. Mathematics and Financial Economics , volume=. 2024 , publisher=

2024
[27]

The Annals of Probability , number =

Ren. The Annals of Probability , number =. 2016 , doi =

2016

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.

[1] [1]

2024 43rd Chinese Control Conference (CCC) , pages=

Direct approach of linear-quadratic Stackelberg mean field games of backward-forward stochastic systems , author=. 2024 43rd Chinese Control Conference (CCC) , pages=. 2024 , organization=

2024

[2] [2]

SIAM journal on Control and Optimization , volume=

Linear-quadratic optimal control problems for mean-field stochastic differential equations , author=. SIAM journal on Control and Optimization , volume=. 2013 , publisher=

2013

[3] [3]

2020 39th Chinese Control Conference (CCC) , pages=

A linear quadratic Stackelberg game of backward stochastic differential equations with partial information , author=. 2020 39th Chinese Control Conference (CCC) , pages=. 2020 , organization=

2020

[4] [4]

Advances in Difference Equations , volume=

Backward-forward linear-quadratic mean-field Stackelberg games , author=. Advances in Difference Equations , volume=. 2021 , publisher=

2021

[5] [5]

General theory , author=

Statistics of random processes: I. General theory , author=. 2013 , publisher=

2013

[6] [6]

2018 , publisher=

An introduction to optimal control of FBSDE with incomplete information , author=. 2018 , publisher=

2018

[7] [7]

arXiv preprint arXiv:2412.11433 , year=

Mean field games of major-minor agents with recursive functionals , author=. arXiv preprint arXiv:2412.11433 , year=

arXiv

[8] [8]

Mathematics of Operations Research , volume=

Optimization under rational expectations: A framework of fully coupled forward-backward stochastic linear quadratic systems , author=. Mathematics of Operations Research , volume=. 2023 , publisher=

2023