Pith. sign in

REVIEW 6 minor 35 references

The paper proves that an infinite-horizon linear-quadratic control problem with regime-switching jumps is closed-loop solvable exactly when a coupled algebraic Riccati system has a static stabilizing solution.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

The claimed closed-loop solvability theorem for Markov regime-switching jump-diffusion SLQ problems is invalid because the jump terms in the stability and Riccati equations are expanded incorrectly.

T0 review reviewed 2026-07-31 challenge →

load-bearing objection The reader's rejection is based on a compensator error; the paper's jump expansion is correct and the central equivalence appears sound.

arxiv 2607.24004 v1 pith:6ZHBCNOW submitted 2026-07-27 math.OC

Closed-loop solvability of infinite-horizon stochastic linear-quadratic problem for Markov regime-switching jump-diffusion system

Kai Ding , Fan Wu , Jie Xiong , Xinyue Zhang This is my paper
classification math.OC MSC 93E0393E15
keywords SLQ problemMarkovian jumpsInfinite horizonL2-stabilityClosed-loop solvabilityregime-switchingjump-diffusioncoupled algebraic Riccati equations
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

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 extends infinite-horizon stochastic linear-quadratic (SLQ) control to systems whose state jumps at the same moments the driving Markov chain switches regimes. Its central claim is a complete characterization: the problem is closed-loop solvable if and only if a system of coupled algebraic Riccati equations (CAREs) admits a static stabilizing solution, and then the optimal control is a linear function of the current state, with value function quadratic in the state. The authors also apply this result to a lifetime wealth tracking problem, deriving an explicit optimal investment strategy that becomes more conservative when a switch could cause a price crash. The proof routes through an L^2-stability criterion for the uncontrolled jump-diffusion, which is the load-bearing step of the argument.

Core claim

The core discovery is Theorem 3.1: Problem (M-SLQ)_∞ is closed-loop solvable if and only if the CAREs (18) admit a static stabilizing solution P in D(S^n). When the solution exists, the optimal closed-loop strategy is given by Θ(i) = −N(P,i)† L(P,i)⊤ + (I − N(P,i)† N(P,i)) Π(i) for an arbitrary Π chosen so the closed loop is stable, and the value function is V(x,i) = ⟨P(i)x, x⟩. This converts an infinite-dimensional optimization problem into a finite system of coupled algebraic equations whose solution directly yields both the optimal feedback and the cost. The authors further show the CAREs have at most one static stabilizing solution, since the value function is unique.

What carries the argument

The central object is the system of coupled algebraic Riccati equations (18), built from the operators M(P,i), L(P,i), N(P,i) in (14). These operators are the coefficients of the quadratic form that appears after applying Itô's rule to ⟨P(α_t)X,X⟩ and completing the square; the stability requirement enters through Proposition 2.1, an L²-stability criterion for the uncontrolled Markov-jump linear system (5) stated in terms of the Lyapunov inequality (6). The pseudo-inverse (Moore–Penrose) and the extended Schur lemma allow the Riccati equations to handle indefinite cost weights, and the static stabilizing solution P, together with the free parameter Π, produces the closed-loop feedback.

Load-bearing premise

Everything hinges on how the quadratic form ⟨P X, X⟩ expands at the instant the regime switches; if the wrong matrix multiplies the jump in that expansion, the stability test that feeds the entire characterization fails.

What would settle it

Two-regime scalar counterexample: set A(i)=C(i)=0, switching rates π_12=π_21=1, and jump amplitudes E_2(1)=E_1(2)=−0.5. Every switch halves the state, so E[X(t)^2]=x^2 e^{−0.75t} and E∫_0^∞ X² dt = x²/0.75 < ∞, making the system L²-stable. Yet condition (6) reduces to 0.25P(2)<0 in regime 1 and 0.25P(1)<0 in regime 2, which no non-negative P can satisfy. Checking this two-mode instance settles whether Proposition 2.1 holds.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Closed-loop solvability of the infinite-horizon problem is fully characterized by a finite system of algebraic Riccati equations, so checking solvability becomes a numerical problem in dimension L·n(n+1)/2.
  • Whenever the CAREs have a static stabilizing solution, an optimal control exists in linear state-feedback form, and the optimal cost is quadratic with the same P that solves the CAREs.
  • Indefinite cost matrices Q and R are allowed; the pseudo-inverse formulation covers singular control weights.
  • For the lifetime wealth tracking application, the optimal investment is an affine feedback on the tracking error, and the presence of regime-switch price jumps makes the optimal risky position more conservative in a bull regime.
  • The model unifies the classical Markov-modulated diffusion case (no jumps) with Poisson-jump diffusions by letting the jump intensities be the transition rates of the chain.

Where Pith is reading between the lines

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

  • Editorial inference: the Itô expansion of ⟨P(α_t)X,X⟩ at a switch should give cross terms P(j)E_j(i)+E_j(i)⊤P(j), not (P(j)-P(i))E_j(i)+E_j(i)⊤(P(j)-P(i)); the pre-switch matrix P(i) should not multiply the jump. Under the corrected expansion, the Lyapunov inequality and the CAREs would take a different form, and Proposition 2.1 would need revision.
  • Editorial inference: a two-regime scalar test with zero drift, zero volatility, symmetric switching rates, and jumps that halve the state at every switch is L²-stable but violates condition (6), so the stated stability criterion appears to be false as written; consequently the Riccati characterization in Theorem 3.1, which relies on it, would not follow as stated.
  • Editorial inference: if the cross-term error is repaired, the same completing-the-square proof strategy may still work, yielding a corrected system of CAREs; recomputing the paper's two-regime numerical example with the corrected equations would test whether the qualitative conclusion about conservative investment at regime switches survives.
  • Editorial inference: the affine wealth-tracking result in Section 4 inherits the stability criterion; until the Lyapunov step is fixed, the feedback formula (38) and the reported sensitivity to jump risk should be treated as provisional.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

0 major / 6 minor

Summary. The paper studies infinite-horizon stochastic linear-quadratic control for linear systems modulated by a Markov chain whose state jumps synchronously with regime switches, with the jump term driven by compensated counting martingales. It characterizes L^2-stability by a coupled Lyapunov inequality (Proposition 2.1), proves that closed-loop solvability of the indefinite SLQ problem is equivalent to the existence of a static stabilizing solution of a coupled algebraic Riccati equation (Theorem 3.1), and gives the feedback representation of the optimal strategy. An application to a lifetime wealth tracking problem is developed and illustrated numerically.

Significance. The model is a natural and useful extension of Markov-modulated SLQ control: regime switches now cause endogenous jumps in the state, not merely changes of coefficients. The main theorem is the first characterization of closed-loop solvability for this class and is likely to be of interest to the control community. The paper is self-contained; the Lyapunov and Riccati conditions are derived explicitly and no parameters are fitted to data. I carefully checked the Itô-type expansions in Eqs. (6), (8), and (14): they are consistent with the compensated-jump formulation, because the dÑ_j term induces a compensating drift -Σλ_j(E_jX+F_ju)dt, and the cross-terms (P(j)-P(i))E_j and (P(j)-P(i))F_j are correct. A counterexample based on the raw jump expansion is not applicable to this model.

minor comments (6)
  1. [Section 1, Eq. (1)] The SDE (1) is written with compensated martingales dÑ_j. This implies an additional drift -Σ_j λ_j(E_jX+F_ju)dt between switches. The pathwise jump at a switch is indeed ΔX=E_jX+F_ju, but the infinitesimal generator contains the extra -f_xΔ term. The paper should state this explicitly; the current text describes only the pathwise jump and invites the common mistake of omitting the compensator drift.
  2. [Lemma 3.1, after Eq. (16)] The equality lim_{T→∞}E⟨P(α_T)X(T),X(T)⟩=0 is used but not justified. For general L^2 processes this is false; for linear SDEs with L^2 inputs it is true (via the differential equation for the second moment and integrability of the right-hand side), but a proof or reference should be supplied.
  3. [Theorem 3.1, necessity proof] The dynamic programming principle is invoked without stating its hypotheses or giving a reference. Since the state process has jumps and the cost coefficients are indefinite, please add a standard DPP statement or a specific citation.
  4. [Definition of U_ad] The admissible set does not require the cost functional to be finite. For indefinite Q,R this could in principle allow J=-∞. It would be cleaner to require the cost to be well-defined or to add a sentence noting that under closed-loop solvability no admissible control gives -∞.
  5. [Proposition 2.1 and Eq. (6)] The matrix inequality in (6) should explicitly say 'negative definite' (or use ≺ 0), and the sentence 'P(τ,i) is increasing in τ' should specify the Loewner order.
  6. [General] Minor typographical and formatting issues: the problem label '(M-SLQ)∞' appears in inconsistent forms; the caption of Figure 2 says 'Riccati solutions' while plotting P(1),P(2); consider adding a closing remark that the jump matrix notation E_j(i) denotes the jump when switching from i to j.

Circularity Check

0 steps flagged

No significant circularity; the central characterization is derived self-containedly, with only a peripheral self-citation for a standard transformation.

full rationale

The main derivation chain is self-contained. Proposition 2.1 is proved internally through coupled Lyapunov-type ODEs and a monotone-convergence argument, so the L2-stability criterion (6) is not assumed as an input. Theorem 3.1 is obtained by applying Itô's formula to ⟨P(αt)X,X⟩, completing the square with the pseudoinverse, and using the CAREs (18); the value function V(x,i)=⟨P(i)x,x⟩ is derived, not posited as the definition of P. The CAREs are defined independently of the value function in (14), and the equivalence in Theorem 3.1 is a genuine if-and-only-if statement rather than a renaming of a fitted quantity. The numerical section solves the paper's own CAREs and does not present fitted parameters as predictions of external data. The only operational self-citation is the 'linear substitution method introduced in Wu et al. [28]' used in Section 4; the substitution is explicitly displayed in (32)-(34), is a standard exponential-discount transformation, and does not carry the burden of the main theorem. Thus no circular step is exhibited; the relevant caveats about the paper concern mathematical rigor or correctness of the jump generator expansion, not circularity.

Axiom & Free-Parameter Ledger

1 free parameters · 7 axioms · 1 invented entities

The mathematical claim depends on the standard Itô/generator calculus, the extended Schur lemma, and assumptions (H1), (H1)', irreducibility. The theory introduces no fitted constants; the only hand-chosen numbers are the illustrative market parameters in §5. The critical failure is an axiom-level error: the jump cross-term convention in (6)/(14) is ad hoc and contradicts the direct expansion of the SDE (1).

free parameters (1)
  • Section 5 market parameters and penalty weight = r=(0.03,0.01), μ=(0.12,−0.04), σ=(0.12,0.20), ρ=(0.05,0.03), γ_2(1)=−0.18, γ_1(2)=0.05, π_12=0.2, π_21=0.1, λ=0.02
    Chosen by hand for the numerical illustration; they determine P(1), P(2), Θ̂ but are not part of the mathematical theorem.
axioms (7)
  • standard math Itô formula / generator for Markov-modulated jump-diffusions: at an i→j switch, Δ⟨P(α)X,X⟩ = ⟨P(j)X_new, X_new⟩ − ⟨P(i)X, X⟩ with X_new = (I+E_j)X + F_j u
    Used throughout (Lemma 3.1, Prop 2.1). This is where the paper's (P(j)−P(i))-weighted cross-terms depart from the correct P(j)-weighted ones (Eqs. (6),(14)).
  • standard math Extended Schur lemma with pseudoinverses (Albert [3], Penrose [17])
    Bridge between the matrix inequality and the CAREs (18); cited.
  • domain assumption (H1) closed-loop strategy set H_α is nonempty
    Assumed before Theorem 3.1; guarantees U_ad nonempty (Remark 3.1).
  • domain assumption Markov chain α is irreducible with finite state space S
    Stated in §1; used implicitly for existence of solutions and the generator structure.
  • domain assumption (H1)' ρ(i) > r(i) in the wealth-tracking application
    Introduced in §4 so that (33) is L²-stable via Remark 2.1 — which itself rests on the disputed criterion (6).
  • domain assumption Dynamic programming principle holds for Problem (M-SLQ)_∞ with the candidate value ⟨P(i)x,x⟩
    Invoked in the necessity proof of Theorem 3.1 without proof or citation for this infinite-horizon indefinite-jump setting.
  • ad hoc to paper The jump contribution to the Itô expansion of ⟨P(α)X,X⟩ has cross-terms (P(j)−P(i))E_j + E_j⊤(P(j)−P(i)) and (P(j)−P(i))F_j (Eqs. (6),(14))
    Never derived; contradicts the direct expansion of (1). Under this 'axiom', Prop 2.1 is false (0.25P(2)<0 for a stable halving system).
invented entities (1)
  • Regime-switch-synchronous state jumps (ΔX = E_j(i)X(t−) + F_j(i)u(t) when α switches i→j) independent evidence
    purpose: Model instantaneous shocks (e.g., price crashes) at regime transitions, beyond pure coefficient switching
    The jump model is the paper's main contribution and is falsifiable outside the paper (e.g., by calibrating asset-price jumps at regime-switch dates); it is not an ad hoc artifact, though the paper's equations for it are wrong.

reviewed 2026-07-31 · how reviews work

0 comments
Cite this review

Pith. "Pith review of Closed-loop solvability of infinite-horizon stochastic linear-quadratic problem for Markov regime-switching jump-diffusion system." pith.science (2026). https://pith.science/paper/6ZHBCNOW

@misc{pith2026260724004,
  author       = {Pith},
  title        = {Pith review of: Closed-loop solvability of infinite-horizon stochastic linear-quadratic problem for Markov regime-switching jump-diffusion system},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6ZHBCNOW}},
  note         = {Machine review of arXiv:2607.24004}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

This paper investigates a class of stochastic linear-quadratic (SLQ) control problems over an infinite horizon for Markov regime-switching jump-diffusion systems. Unlike classical diffusion models modulated by a Markov chain, we assume that the state process undergoes abrupt jumps that are synchronous with the regime switches of the Markov chain. In contrast to conventional Poisson jump-diffusion models, the jumps in the state process are entirely induced by the state transitions of the Markov chain, which can be interpreted as losses or gains of state process incurred during regime changes. Under this formulation, we thoroughly discuss the closed-loop solvability of the SLQ control problem and provide a feedback representation of the optimal control via the stabilizing solution of a system of coupled algebraic Riccati equations (CAREs). Finally, we further apply our results to a lifetime wealth tracking problem and derive the corresponding optimal investment strategy.

Figures

Figures reproduced from arXiv: 2607.24004 by Fan Wu, Jie Xiong, Kai Ding, Xinyue Zhang.

Figure 1
Figure 1. Figure 1: Sample paths of the optimal wealth X∗ (t), target G(t), and investment control u ∗ (t). 20 [PITH_FULL_IMAGE:figures/full_fig_p020_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Sensitivity of the stabilizing solution with respect to the control weight [PITH_FULL_IMAGE:figures/full_fig_p021_2.png] view at source ↗

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

35 extracted references · 1 linked inside Pith

  1. [1]

    Linear matrix inequalities, Riccati equations, and indefinite stochastic linear quadratic controls.IEEE Transactions on Automatic Control, 45(6):1131–1143, 2000

    Mustapha Ait Rami and Xun Yu Zhou. Linear matrix inequalities, Riccati equations, and indefinite stochastic linear quadratic controls.IEEE Transactions on Automatic Control, 45(6):1131–1143, 2000

    2000

  2. [2]

    Mustapha Ait Rami, Xun Yu Zhou, and J.B. Moore. Well-posedness and attainability of indefinite stochas- tic linear quadratic control in infinite time horizon.Systems & Control Letters, 41:123–133, 2000

    2000

  3. [3]

    Conditions for positive and nonnegative definiteness in terms of pseudoinverses.SIAM Journal on Applied Mathematics, 17(2):434–440, 1969

    Arthur Albert. Conditions for positive and nonnegative definiteness in terms of pseudoinverses.SIAM Journal on Applied Mathematics, 17(2):434–440, 1969. 21

    1969

  4. [4]

    Markovian transition counting processes: An alternative to Markov modulated Poisson processes

    Azam Asanjarani, Sophie Hautphenne, and Yoni Nazarathy. Markovian transition counting processes: An alternative to Markov modulated Poisson processes. arXiv:2109.00713, 2021

    Pith/arXiv arXiv 2021

  5. [5]

    Yuzhu Bai, Hui Jie Sun, and Ai Guo Wu. Linear quadratic optimal control for a class of continuous-time nonhomogeneous Markovian jump linear systems in infinite time horizon.Journal of the Franklin Institute, 357(14):9733–9760, 2020

    2020

  6. [6]

    Linear quadratic optimal stochastic control with random coefficients.SIAM Journal on Control and Optimization, 14(3):419–444, 1976

    Jean Michel Bismut. Linear quadratic optimal stochastic control with random coefficients.SIAM Journal on Control and Optimization, 14(3):419–444, 1976

    1976

  7. [7]

    Tian Chen and Yunxiao Jia. A general maximum principle for progressive optimal control of partially observed mean-field stochastic system with Markov chain.ESAIM: Control, Optimisation and Calculus of Variations, 29:58, 2023

    2023

  8. [8]

    Stochastic linear-quadratic control with conic control constraints on an infinite time horizon.SIAM Journal on Control and Optimization, 43(3):1120–1150, 2004

    Xi Chen and Xun Yu Zhou. Stochastic linear-quadratic control with conic control constraints on an infinite time horizon.SIAM Journal on Control and Optimization, 43(3):1120–1150, 2004

    2004

  9. [9]

    The stochastic maximum principle for relaxed control problem with regime-switching.Systems & Control Letters, 169:105391, 2022

    Yinggu Chen, Tianyang Nie, and Zhen Wu. The stochastic maximum principle for relaxed control problem with regime-switching.Systems & Control Letters, 169:105391, 2022

    2022

  10. [10]

    Constrained stochastic LQ control on infinite time horizon with regime switching.ESAIM: Control, Optimisation and Calculus of Variations, 28:Paper No

    Ying Hu, Xiaomin Shi, and Zuo Quan Xu. Constrained stochastic LQ control on infinite time horizon with regime switching.ESAIM: Control, Optimisation and Calculus of Variations, 28:Paper No. 5, 2022

    2022

  11. [11]

    A linear-quadratic optimal control problem for mean-field stochastic differential equations in infinite horizon.Mathematical Control and Related Fields, 5(1):97–139, 2015

    Jianhui Huang, Xun Li, and Jiongmin Yong. A linear-quadratic optimal control problem for mean-field stochastic differential equations in infinite horizon.Mathematical Control and Related Fields, 5(1):97–139, 2015

    2015

  12. [12]

    Stochastic linear quadratic optimal control problem: A reinforcement learning method.IEEE Transactions on Automatic Control, 67(9):5009–5016, 2022

    Na Li, Xun Li, Jing Peng, and Zuo Quan Xu. Stochastic linear quadratic optimal control problem: A reinforcement learning method.IEEE Transactions on Automatic Control, 67(9):5009–5016, 2022

    2022

  13. [13]

    Indefinite stochastic linear quadratic control with Marko- vian jumps in infinite time horizon.Journal of Global Optimization, 27:149–175, 2003

    Xun Li, Xun Yu Zhou, and Mustapha Ait Rami. Indefinite stochastic linear quadratic control with Marko- vian jumps in infinite time horizon.Journal of Global Optimization, 27:149–175, 2003

    2003

  14. [14]

    Mean-field linear-quadratic stochastic differential games in an infinite horizon.ESAIM: Control, Optimisation and Calculus of Variations, 27(81):40, 2021

    Xun Li, Jingtao Shi, and Jiongmin Yong. Mean-field linear-quadratic stochastic differential games in an infinite horizon.ESAIM: Control, Optimisation and Calculus of Variations, 27(81):40, 2021

    2021

  15. [15]

    Stochastic maximum principle for forward-backward regime switching jump diffusion systems and applications to finance.Chinese Annals of Mathematics, Series B, 39(5):773–790, 2018

    Siyu Lv and Zhen Wu. Stochastic maximum principle for forward-backward regime switching jump diffusion systems and applications to finance.Chinese Annals of Mathematics, Series B, 39(5):773–790, 2018

    2018

  16. [16]

    Nguyen, George Yin, and Dung T

    Son L. Nguyen, George Yin, and Dung T. Nguyen. A general stochastic maximum principle for mean-field controls with regime switching.Applied Mathematics & Optimization, 84(3):3255–3294, 2021

    2021

  17. [17]

    A generalized inverse for matrices.Mathematical Proceedings of the Cambridge Philosophical Society, 51(3):406–413, 1955

    Roger Penrose. A generalized inverse for matrices.Mathematical Proceedings of the Cambridge Philosophical Society, 51(3):406–413, 1955

    1955

  18. [18]

    Constrained stochastic LQ optimal control problem with random coefficients on infinite time horizon.Applied Mathematics & Optimization, 83(2):1005–1023, 2021

    Jiangyan Pu and Qi Zhang. Constrained stochastic LQ optimal control problem with random coefficients on infinite time horizon.Applied Mathematics & Optimization, 83(2):1005–1023, 2021. 22

    2021

  19. [19]

    Optimal mean-variance portfolio selection under regime-switching-induced stock price shocks.Systems & Control Letters, 204:106200, 2025

    Xiaomin Shi and Zuo Quan Xu. Optimal mean-variance portfolio selection under regime-switching-induced stock price shocks.Systems & Control Letters, 204:106200, 2025

    2025

  20. [20]

    Yizhan Shu, Chenyu Yu, and John M. Mulvey. Downside risk reduction using regime-switching signals: A statistical jump model approach.Journal of Asset Management, 25(5):493–507, 2024

    2024

  21. [21]

    Yuanzhuo Song and Zhen Wu. A general maximum principle for progressive optimal stochastic control problems with Markov regime-switching.ESAIM: Control, Optimisation and Calculus of Variations, 28 (61):19, 2022

    2022

  22. [22]

    Stochastic linear quadratic optimal control problems in infinite horizon

    Jingrui Sun and Jiongmin Yong. Stochastic linear quadratic optimal control problems in infinite horizon. Applied Mathematics & Optimization, 78:145–183, 2018

    2018

  23. [23]

    Zhongyang Sun, Junyi Guo, and Xin Zhang. Maximum principle for Markov regime-switching forward– backward stochastic control system with jumps and relation to dynamic programming.Journal of Opti- mization Theory and Applications, 176:319–350, 2018

    2018

  24. [24]

    Zhongyang Sun, Isabelle Kemajou-Brown, and Olivier Menoukeu-Pamen. A risk-sensitive maximum princi- ple for a Markov regime-switching jump-diffusion system and applications.ESAIM: Control, Optimisation and Calculus of Variations, 24(3):985–1013, 2018

    2018

  25. [25]

    Shanjian Tang. General linear quadratic optimal stochastic control problems with random coefficients: Linear stochastic Hamilton systems and backward stochastic Riccati equations.SIAM Journal on Control and Optimization, 42(1):53–75, 2003

    2003

  26. [26]

    Mean-variance portfolio selection under a non-Markovian regime-switching model: Time-consistent solutions.SIAM Journal on Control and Optimization, 57(5): 3249–3271, 2019

    Tianxiao Wang, Zhuo Jin, and Jiaqin Wei. Mean-variance portfolio selection under a non-Markovian regime-switching model: Time-consistent solutions.SIAM Journal on Control and Optimization, 57(5): 3249–3271, 2019

    2019

  27. [27]

    Qingmeng Wei and Zhiyong Yu. Infinite horizon forward-backward SDEs and open-loop optimal controls for stochastic linear-quadratic problems with random coefficients.SIAM Journal on Control and Optimization, 59(4):2594–2623, 2021

    2021

  28. [28]

    Stochastic linear quadratic optimal control problems with regime- switching jumps in infinite horizon.SIAM Journal on Control and Optimization, 63(2):852–891, 2025

    Fan Wu, Xun Li, and Xin Zhang. Stochastic linear quadratic optimal control problems with regime- switching jumps in infinite horizon.SIAM Journal on Control and Optimization, 63(2):852–891, 2025

    2025

  29. [29]

    Yao, Shuzhong Zhang, and Xun Yu Zhou

    David D. Yao, Shuzhong Zhang, and Xun Yu Zhou. Stochastic linear-quadratic control via primal-dual semidefinite programming.SIAM Review, 46(1):87–111, 2004

    2004

  30. [30]

    Yao, Shuzhong Zhang, and Xun Yu Zhou

    David D. Yao, Shuzhong Zhang, and Xun Yu Zhou. Tracking a financial benchmark using a few assets. Operations Research, 54(2):232–246, 2006

    2006

  31. [31]

    Linear-quadratic optimal control problems for mean-field stochastic differential equations

    Jiongmin Yong. Linear-quadratic optimal control problems for mean-field stochastic differential equations. SIAM Journal on Control and Optimization, 51(4):2809–2838, 2013

    2013

  32. [32]

    Fu Zhang, Yuchao Dong, and Qingxin Meng. Backward stochastic Riccati equation with jumps associated with stochastic linear quadratic optimal control with jumps and random coefficients.SIAM Journal on Control and Optimization, 58(1):393–424, 2020. 23

    2020

  33. [33]

    Portfolio selection in the enlarged Markovian regime- switching market.SIAM Journal on Control and Optimization, 48(5):3368–3388, 2010

    Xin Zhang, Tak Kuen Siu, and Qingbin Meng. Portfolio selection in the enlarged Markovian regime- switching market.SIAM Journal on Control and Optimization, 48(5):3368–3388, 2010

    2010

  34. [34]

    Elliott, Tak Kuen Siu, and Junyi Guo

    Xin Zhang, Robert J. Elliott, Tak Kuen Siu, and Junyi Guo. Markovian regime-switching market completion using additional Markov jump assets.IMA Journal of Management Mathematics, 23(3):283–305, 2011

    2011

  35. [35]

    Elliott, and Tak Kuen Siu

    Xin Zhang, Robert J. Elliott, and Tak Kuen Siu. A stochastic maximum principle for a Markov regime- switching jump-diffusion model and its application to finance.SIAM Journal on Control and Optimization, 50(2):964–990, 2012. 24

    2012

This paper was first reviewed by deepseek-v4-flash on July 31, 2026.