Pith. sign in

REVIEW 5 major objections 5 minor 14 references

Modelling Financial Market Imperfection Using Open Quantum Systems

T0 review · 5 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A market's off-diagonal entries are conserved, and each orbit has a unique Toeplitz attractor, so the long-run state depends on initial coherences.

desk verdict A genuinely new orbit/attractor framework for quantum-finance market imprecision, but the load-bearing proofs don't hold up as written. read the letter →

arxiv 2505.01284 v1 pith:IQTHF5FA submitted 2025-05-02 q-fin.MF q-fin.GN

Will Hicks This is my paper · ORCID
classification q-fin.MFq-fin.GN MSC 91G8081S2215B05
keywords quantumfinanceopensystemsvonNeumannentropyergodicityself-referentialmarketendogenouspricechangesnon-classicaldiffusionimprecision
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 extends an open-quantum-systems model of financial markets to characterise how market imperfections—illiquid trades, non-standard contracts, imperfect execution—enter the price dynamics alongside ordinary information-driven uncertainty. The central aim is to show that the market state, a reduced density matrix, evolves inside invariant orbits defined by conserved sums of off-diagonal entries, and that each orbit contains a unique stationary attractor of Toeplitz form. The author then distinguishes classical diffusion, which keeps a diagonal density matrix diagonal, from non-classical diffusion, which generates off-diagonal correlations, and argues that both eventually reach the same maximum-entropy state, but classical systems get there faster. If correct, the framework gives a mathematical language for representing market imprecision as a source of uncertainty independent of information arrival, and for separating type I (observable or state diagonal) from type II (both non-diagonal) non-classical regimes.

What carries the argument

The central object is the family of invariant orbits $D_\epsilon(\mathcal{H})$ defined by conserved off-diagonal sums (eq. 19), together with the Toeplitz stationary states $T_\epsilon$ (eq. 21). These play the role of fixed manifolds: the dynamics never leaves the orbit set by the initial off-diagonal sums, and within each orbit the Toeplitz matrix is the unique limiting point, so all memory of the initial state that survives at long times is encoded in the constants $\epsilon_j$. The engine of the argument is the master equation (7), written in Gorini–Kossakowski–Sudarshan/Lindblad form with jump operators $A_u$ and $A_d$, whose structure—sums of shifts along diagonals—forces the orbit conservation and makes the Toeplitz matrices stationary.

What would settle it

Take an off-diagonal perturbation $M_\delta$ consisting of a single nonzero element just off the diagonal, say $M_{i,i+1}=1$, and compute $\|\phi(M_\delta)\|$ under the map of eq. (7) for $\sigma>0$, $\nu_u=\nu_d=0$, using the same $\delta t$ as the paper. If for any such $i$ and small $\delta t>0$ the inequality $\|\phi(M_\delta)\| < \|M_\delta\|$ fails, then Proposition 4.6's contraction argument collapses, and the claimed convergence of every orbit to its Toeplitz state needs another proof.

Watch

Extended reading notes

Core claim

The paper's core discovery is that the Lindblad-type master equation (7) for the reduced market density matrix $\rho_{\mathrm{mkt}}(t)$ partitions the state space into invariant orbits $D_\epsilon(\mathcal{H})$, each fixed by the conserved off-diagonal sums $\sum_i \rho_{i,i+j}(t)=\epsilon_j$ (eq. 19). Within each orbit, the unique stationary point is a Toeplitz matrix $T_\epsilon$ (eq. 21), which Proposition 4.6 asserts is a stable attractor in the Frobenius norm. Equivalently, the off-diagonal pattern of the initial state fixes the long-run equilibrium, so two markets that differ only in their initial coherences converge to different limiting states. The author then shows that the dynamics is classical—meaning a diagonal state stays diagonal—if and only if $\nu_u^2=\nu_d^2=0$, and that for non-classical parameters the approach to the maximum-entropy state is slower, as demonstrated by numerical simulation with an entropy-based market-precision metric $P_{\mathrm{ent}}$.

Load-bearing premise

The proof of the attractor's uniqueness assumes every small perturbation shrinks in Frobenius norm, but the calculation only checks this for one type of entry and does not verify it for off-diagonal entries, so the long-time convergence result depends on that unverified contraction.

Editorial extensions

If this is right

  • If the central claim is right, illiquid trades and imperfect trading mechanisms can be represented by non-classical diffusion parameters ($\nu_u,\nu_d\neq 0$), giving price uncertainty from market imprecision on top of information entropy.
  • Orbit invariance means the off-diagonal sums $\epsilon_j$ are conserved, so the long-run market state is determined by the initial coherence pattern, not just the initial probabilities.
  • Classical and non-classical dynamics converge to the same maximum-entropy state, but classical convergence is faster, so any observed slow convergence to a uniform distribution in a market could indicate non-classical effects.
  • Type I non-classical systems (one of observable or state diagonal) still allow interpretation of diagonal elements as probabilities, while type II systems do not—so the practical interpretation of measured prices depends on which regime holds.

Reading between the lines

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

  • The orbit decomposition suggests an empirical signature: if real markets displayed slow convergence to maximum entropy relative to a classical benchmark, that would be evidence of non-classical (coherence-preserving) modes—one could test this by estimating the off-diagonal sums $\epsilon_j$ from price data and checking whether they remain conserved over time.
  • The distinction between type I and type II non-classical systems could be mapped to practical market microstructure: type I corresponds to a liquid market where the quoted price operator is aligned with the market state, while type II corresponds to situations where the trade itself changes the basis—an analogy to illiquidity that suggests a calibration route via the trade-size distribution.
  • If the Toeplitz attractor is the long-run equilibrium for each orbit, then a market that starts with coherences (e.g., from a non-standard contract) will never relax to the fully mixed state; the residual off-diagonal structure would manifest as persistent price misalignment, measurable as a non-zero $P_{\mathrm{ent}}$ limit.
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.

Referee Report

5 major / 5 minor

Summary. The paper applies open quantum systems to financial market modelling, proposing a Markovian master equation (equation 7) for the market's reduced density matrix under exogenous news events. Building on the author's earlier work, it characterizes the long-time evolution via conserved 'diffusion orbits' D_epsilon(H) defined by off-diagonal sums (equation 19), claims that each orbit contains a unique stationary Toeplitz attractor T_epsilon (equation 21), and distinguishes classical from non-classical diffusion. It then introduces type I and type II non-classical systems, proposes two metrics for 'market precision' (P_var and P_ent), and presents numerical simulations suggesting that classical diffusion reaches the maximum-entropy state faster than non-classical diffusion. The paper concludes that non-classical modes can represent illiquid trades and imperfect trading mechanisms.

Significance. If the central claims were rigorously established, the paper would offer a novel mathematical bridge between quantum information concepts and financial microstructure, with explicit orbit structure, entropy monotonicity, and a concrete distinction between classical and non-classical uncertainty. The conceptual proposals—conserved off-diagonal sums as invariants, non-diagonal price operators for market imprecision, and the two precision metrics—are original and potentially useful. However, the load-bearing mathematical results are not proven in the submitted form: the contraction proof in Proposition 4.6 is invalid, the complete-positivity condition in Proposition 3.2 appears incorrect and is violated by the paper's own simulations, and the entropy monotonicity proof in Proposition 4.1 has gaps. The paper also lacks external data or independent benchmarks, so the 'classical converges faster' claim rests entirely on simulations of the author's own model. Strong conceptual novelty, but the technical foundation as written is not sound.

major comments (5)
  1. [4.2, Proposition 4.6] The proof of contraction is invalid. In the step 'by the triangle inequality: ||phi(M_delta)|| <= sum_{i,j} ||phi(M_ij |f_i><f_i|)||', the off-diagonal basis elements |f_i><f_j| are replaced with diagonal elements |f_i><f_i|; this is not what the triangle inequality gives, and it drops exactly the off-diagonal directions that define the orbits D_epsilon. Even if each diagonal term were a strict contraction, the bound would be of the form (1 - c delta_t) sum |M_ij|, which is larger than ||M_delta|| because ||M_delta|| <= sum |M_ij|. The subsequent line ||phi(M_ij |f_i><f_i|)||^2 = 1 - 4 sigma^2 delta_t + O(delta_t^2) <= M_ij^2 conflates a scalar amplitude M_ij with a norm and is dimensionally inconsistent. Thus the proof does not establish the Frobenius-norm contraction required by Definition 4.5, and the claim that T_epsilon is a stable attractor and the unique limit point in each orbit is unsupported.
  2. [3.2, Proposition 3.2] The complete-positivity condition nu_u^2 + nu_d^2 <= sigma^2 appears incorrect. Writing equation 7 in the GKS form with F_1 = A_u, F_2 = A_d and c_11 = c_22 = sigma^2, c_12 = -nu_u^2, c_21 = -nu_d^2, the coefficient matrix C = [[sigma^2, -nu_u^2], [-nu_d^2, sigma^2]] is positive semidefinite when sigma^4 >= nu_u^2 nu_d^2, not when the sum of squares is bounded by sigma^2. Moreover, the simulation in Section 6 uses sigma = 0.4, nu_u = nu_d = 0.36, for which nu_u^2 + nu_d^2 = 0.2592 > sigma^2 = 0.16; the paper does not acknowledge that its own simulations are outside the regime identified by Proposition 3.2, nor does it explain whether non-CP evolution is still intended to be physical.
  3. [4, Proposition 4.1] The proof that von Neumann entropy is non-decreasing assumes that I/N is a stationary state of equation 7 and that the dynamics is completely positive. While the dissipative part of equation 7 leaves I/N invariant by direct calculation, the proof does not address the -Trenv[HI(t), rho_I(0)] term, which can be nonzero outside the special case assumed in Section 4.1. More importantly, the relative-entropy monotonicity inequality used in equation (12) is valid for completely positive maps; as Proposition 3.2's condition is not satisfied in general (and is incorrectly stated), the entropy monotonicity claim is not established for the full equation 7.
  4. [4.3, Proposition 4.8] The proof of Proposition 4.8 equates 'orthogonal' with the property that L_i maps basis vectors to single basis vectors (L_i |f_j> = alpha_k |f_k> followed by relabelling to |f_{k+n}>). This property holds for permutation-type operators, not for general orthogonal matrices; a rotation matrix is orthogonal but maps a basis vector to a superposition. Consequently the 'if' direction of the proposition is false as stated, and the characterization of classical diffusion via orthogonality of the Lindblad operators is not established. The closure of the sets D_j under the dynamics is also asserted rather than proven for the full generator with the specific A_u, A_d.
  5. [4.2, Proposition 4.6 (definition of phi)] The one-step map phi in the proof of Proposition 4.6 is defined using the operators L(rho), L(A_u rho A_u), and L(A_d rho A_d), which come from the N -> infinity equation (9), not from the finite-N equation (7) that the proposition claims to treat. This mismatch between the stated dynamics and the map whose contraction is proved is not explained, so even if the contraction estimate were corrected, it would apply to a different evolution than the one whose stationary points are under discussion.
minor comments (5)
  1. [Throughout] The text consistently writes 'V on-Neumann' instead of 'von Neumann'; this should be corrected throughout.
  2. [6.4, first bullet] The text cites 'proposition 3.1' when referring to the non-classical terms for nu_u, nu_d != 0; the intended reference is likely Proposition 3.2 or Proposition 4.8.
  3. [Equation (26)] The expansion |v_i> = epsilon |f_{i-1}> + sqrt(1 - 2 epsilon^2) |f_j> + epsilon |f_{i+1}> uses j rather than i in the middle term; this appears to be a typo.
  4. [Section 6.1, Definition 6.1] The definition of P_var in equation (29) is unclear: the numerator uses max_i Tr[X^2 rho_i] but the rho_i are defined as the components of rho in its eigenbasis; the connection to the stated requirements (zero for diagonal rho, one for pure states with variance) is not immediately evident and should be spelled out.
  5. [Section 6.4] The claim that 'it therefore follows from proposition 4.4 that the off-diagonal sums remain at zero in both simulations' needs a brief explanation of why the initial Gaussian state, which is diagonal, has zero off-diagonal sums; this is true but not explicitly noted.

Circularity Check

0 steps flagged · score 2.0 of 10

No circularity by construction; the main risk is an invalid contraction proof, not a self-referential reduction.

full rationale

I walked the derivation chain from Equation (7) through Propositions 4.4, 4.6, and the type I/II discussion. Equation (7) is imported from the author's earlier work ([10], [11]) rather than re-derived here, and all later claims are statements about that same equation. That is a self-citation lineage, but it is not a circular reduction: the orbit invariants in Proposition 4.4 and the Toeplitz stationary candidates in Equation (21) are constructed and argued in this paper, and the uniqueness/stability claim is intended to be established by the contraction argument in Proposition 4.6, not simply assumed from the citation. The definitions of D_epsilon and T_epsilon use the same off-diagonal sums, but this is a natural parametrization of the invariant orbits, not an equation that is assumed equal to the conclusion. The simulations in Section 6 use the same model to illustrate the claimed behaviour; they are not calibrated to external data, but they also are not a fitted parameter relabelled as a prediction. I found no step in which a stated prediction is equivalent by construction to a fitted input or to a prior self-cited theorem. The serious problem with Proposition 4.6 is mathematical validity, not circularity: the proof replaces a general perturbation M_delta by diagonal basis elements, applies the triangle inequality in the wrong direction, and treats the scalar M_ij as a norm bound, so the asserted contraction in off-diagonal zero-sum directions is not established. Because the faster-classical claim depends on that contraction, this is a substantive correctness gap; but a flawed proof is not an input-output equivalence, so I do not count it as circularity. Score 2 reflects the minor self-citation lineage of Equation (7) as the underlying dynamical premise, not a demonstrated circular step.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The model rests on the author's previously constructed open quantum system framework, with hand-chosen simulation parameters and standard quantum mechanics axioms. No new physical entities are introduced, but the financial validity of the Hilbert space and interaction Hamiltonian is assumed rather than empirically grounded.

free parameters (5)
  • Environment coupling strength sigma = 0.4 (simulation, Section 6.2)
    Set by hand for the numerical simulation; not derived from any market data.
  • Non-classical diffusion rates nu_u and nu_d = 0.36 each (simulation, Sections 5.4.1 and 6.3)
    Chosen ad hoc to illustrate the non-classical regime; no calibration to financial data.
  • Initial Gaussian width sigma_0 = 0.05 (equation 31)
    Arbitrary choice for the starting price distribution in the simulation.
  • Lattice size N and time step delta_t = N=101, delta_t=0.004 (Section 6.2)
    Chosen for numerical convenience and stability; results may depend on these choices.
  • Random unitary matrix U for type II simulation = not specified
    Used to construct the type II simulation in Section 5.4.1; no seed or explicit matrix is given, so exact reproduction is impossible.
assumptions (5)
  • domain assumption The Markovian master equation (equation 7) derived in [10] and [11] is a valid description of market-environment dynamics.
    The paper takes equation 7 as its starting point without re-deriving the Born-Markov approximations or testing them against financial data.
  • domain assumption Boundary terms can be neglected (N to infinity) in the rewrite of equation 7 as Proposition 3.1, while finite-N statements in Proposition 4.1 and 4.6 are treated as exact.
    This inconsistency is load-bearing: I/N is only approximately stationary in the infinite-volume limit, not for finite N.
  • standard math The standard GKLS form (equation 22, references [5], [8], [12]) characterizes all generators of completely positive trace-preserving semigroups.
    Used in Proposition 3.2 and Proposition 4.8; the complete positivity condition itself is mis-stated in the paper, since it should involve the determinant of the c-matrix, not nu_u^2 + nu_d^2 <= sigma^2.
  • standard math Relative entropy is monotone under completely positive trace-preserving maps (Breuer and Petruccione).
    Used in Proposition 4.1; the monotonicity itself is sound, but the chosen stationary state is not actually stationary.
  • domain assumption Zero drift: the system Hamiltonian on the market space is set to zero for martingale measures.
    Assumed to simplify the model, stated in Section 3.1.1; it excludes any systematic price drift.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Modelling Financial Market Imperfection Using Open Quantum Systems." pith.science (2026). https://pith.science/paper/IQTHF5FA

@misc{pith2026250501284,
  author       = {Pith},
  title        = {Pith review of: Modelling Financial Market Imperfection Using Open Quantum Systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IQTHF5FA}},
  note         = {Machine review of arXiv:2505.01284}
}
read the original abstract

We start with the idea that open quantum systems can be used to represent financial markets by modelling events from the external environment and their impact on the market price. We show how to characterize distinct orbits of the time evolution, and look at the development of the reduced density matrix, that represents the state of the market, over long time frames. In particular we distinguish between classical and non-classical modes of time evolution. We show that whilst both tend to the same set of maximum entropy states, this occurs faster in classical systems, with a knock on effect on the resulting probability distributions. We demonstrate how non-classical modes of time-evolution can be used to incorporate factors such as illiquid trades and imperfect trading mechanisms, and distinguish between different mechanisms of non-classical time evolution.

Figures

Figures reproduced from arXiv: 2505.01284 by the authors.

Figure 1
Figure 1. Reduced density matrix after a single time-step, where we assume the reduced density matrix starts in a Dirac state. The plot shows the non-central matrix elements only. That is it does not depict the large central matrix element of 0.9968. This helps to illustrate the regularity that is due to the simple form for the time-evolution. • Diagonal reduced density matrix: ρmkt(t), representing the case where the uncerta… view at source ↗
Figure 2
Figure 2. Reduced density matrix after a single time-step, where we assume the reduced density matrix starts in a Dirac state, that has first been diagonalized by a random unitary matrix. By diagonalizing the reduced density matrix, the simple form for the time-evolution is no longer evident, and the regular structure of the non-central matrix elements is lost. A f(ρmkt(t)) = 0 for diagonal ρmkt(t). This represents the case w… view at source ↗
Figure 3
Figure 3. The entropy based metric, described by definition 6.2, for simulation 1 and simulation 2 as described above [PITH_FULL_IMAGE:figures/full_fig_p020_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: The excess kurtosis for simulation 1 and simulation 2 as described above 20 [PITH_FULL_IMAGE:figures/full_fig_p020_4.png]
Figure 5
Figure 5. Figure 5: The first four non-zero off-diagonals for simulation 1/2 after 5000 steps: ρi(i+j), i = 1 . . . N for j = 2, 4, 6, 8 [PITH_FULL_IMAGE:figures/full_fig_p022_5.png]
Figure 6
Figure 6. Figure 6: The evolution of the ||D2|| metric defined by equation 32 takes much longer, with the maximum value for ||D2|| being reached after 32K time-steps. 22 [PITH_FULL_IMAGE:figures/full_fig_p022_6.png]
Figure 7
Figure 7. Figure 7: Distance in the Frobenius norm from the maximum entropy state. Theoretical Entropy Limit: • After 5000 time-steps, we separate out simulation 1, which carries on with the non-classical diffusion, from simulation 2, which now diffuses classically. • As per proposition 4…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

14 extracted references · 12 canonical work pages

  1. [1]

    The Role of Information in a Two-Traders Market

    Bagarello, F,: Haven, E,: “The Role of Information in a Two-Traders Market”, Physica A: Statistical Mechanics and its Applications (2014)

    work page 2014

  2. [2]

    Cambridge University Press, 2018, ISBN:978-1- 107-15605-0

    Donier, J,; Bouchaud, J-P,; Bonart, J,; Gould, M,: Trades Quotes and Financial Markets Under The Microscope. Cambridge University Press, 2018, ISBN:978-1- 107-15605-0

    work page 2018

  3. [3]

    Breuer, H.P.; Petruccione, F.: The Theory of Open Quantum Systems , Oxford University Press (2002)

    work page 2002

  4. [4]

    Trade Size, Order Imbalance, and The V olatility-V olume Relation

    Chan, K.; Fong W-M,: “Trade Size, Order Imbalance, and The V olatility-V olume Relation”, Journal Of Financial Economics, 57.2 (2000)

    work page 2000

  5. [5]

    Chru ´sinski, D,; Pascazio, S: A Brief History of the GKLS Equation , arxiv.org/pdf/1710.05993 24 Modelling Financial Market Imperfection Using Open Quantum Systems 25

    arXiv

  6. [6]

    What moves stock prices: Another look

    Cornell, B.: “What moves stock prices: Another look”, The Journal of Portfolio Management (2013) Apr 30;39(3):pp32-8

    work page 2013

  7. [7]

    What Moves Stock Prices

    Cutler, D.M,; Poterba, J.M.; Summers, L.H,: “What Moves Stock Prices”, Journal Of Portfolio Management 15 (3): pp 4-12, doi: 10.3905/jpm.1989.409212

    work page arXiv 1989

  8. [8]

    Gorini, V .; Kossakowski, A.; Sudarshan, E.C.G: Completely Positive Dynamical Semigroups of N-level systems, J. Maths. Phys. 17, 821 (1976)

    work page 1976

Show all 14 references
  1. [9]

    Hall, B.C: Quantum Theory for Mathematicians , Springer 2013, ISBN: 978-1- 4614-7116-5

    2013

  2. [10]

    Hicks, W,: Applications of quantum stochastic calculus and open quantum systems, in the modelling of the financial market.Doctoral (Phd) thesis, Memorial University of Newfoundland (2024)

    2024

  3. [11]

    Information Entropy of the Financial Market: Modelling Random Processes Using Open Quantum Systems

    Hicks, W,: “Information Entropy of the Financial Market: Modelling Random Processes Using Open Quantum Systems”. Quantum Economics and Finance (2024) https://doi.org/10.1177/29767032241279990

    2024 doi

  4. [12]

    Lindblad, G: On the generators of quantum dynamical semigroupsCommun. Math. Phys. 48, 119 (1976)

    1976

  5. [13]

    Press, W.H,; Teukolsky S.A,; Vetterling, W.T,; Flannery, B.P: Numerical Recipes in C++, The Art of Scientific Computing , Second Edition, Cambridge University Press

  6. [14]

    Perez-Garcia, D,; Wolf, M.M,; Petz, D,; Ruskai, M,B,: Contractivity of positive and trace preserving maps underLp norms, arxiv:math-ph/0601063 25

    arXiv

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.