Pith. sign in

REVIEW 4 major objections 3 minor 34 references

A non-local estimator for locally stationary Hawkes processes

T0 review · 4 major / 3 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 On a single observed path of a locally stationary Hawkes process, the ordinary Ogata maximum likelihood estimator remains $\sqrt{T}$-consistent and asymptotically normal, and a Bernstein-polynomial likelihood-ratio test detects any…

desk verdict Useful extension of Ogata MLE to locally stationary Hawkes processes, but the time-invariance test's consistency over C[0,1] is overstated: the Bernstein degree must exceed a g-dependent threshold that the practitioner doesn't know. read the letter →

arxiv 2506.02631 v1 pith:5HFUGJ4C submitted 2025-06-03 math.ST stat.TH

Thomas Deschatre , Pierre Gruet , Antoine Lotz This is my paper
classification math.STstat.TH MSC 62F0362F1260G55
keywords HawkesprocesseslocallystationaryprocessmaximumlikelihoodestimationratiotestBernsteinpolynomialstime-varyingreproductionrateendogeneityintradaypowermarket
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 asks whether the standard maximum likelihood estimator developed for stationary Hawkes processes can be used, without modification, on a Hawkes process whose baseline intensity and reproduction rate vary with rescaled time $t/T$. Its answer is yes in the parametric setting: from a single trajectory over $[0,T]$, the MLE is $\sqrt{T}$-consistent and asymptotically Gaussian, with asymptotic variance given by the integrated Fisher information of the associated stationary processes. It then builds a likelihood-ratio test of the hypothesis that the reproduction rate $g$ is constant on $[0,1]$, proves a $\chi^2(d)$ limit for $\Lambda^d_T$ under the null, and shows the test is consistent against every non-constant continuous $g$ provided the Bernstein degree $d$ exceeds a $g$-dependent threshold. An application to German intraday power-market order flow finds statistically significant session-dependent fluctuations in the endogeneity rate.

What carries the argument

The argument runs through a family of stationary Hawkes processes $N^{x,\infty}$, indexed by rescaled time $x\in[0,1]$, with intensities $\lambda^{x,\infty}_{k,t}(\eta,g)=\Phi_k[\mu(x,\eta)+\sum_l \int_{-\infty}^t g(x)\varphi_{kl}(t-s,\eta)\,dN^{x,\infty}_{l,s}]$, all embedded in the same Poisson noise. Ergodicity of these stationary processes, together with a Riemann-sum scheme borrowed from the univariate theory, converts time-averages along the single non-stationary path into integrals over $x$ of stationary expectations, so the score becomes a Gaussian martingale with covariance $I(\vartheta^*)$. The time-invariance test relies on the Bernstein basis $B_{k,d}(x)=\binom{d}{k}x^k(1-x)^{d-k}$, under which '$g$ is constant' is exactly the statement that all $d+1$ Bernstein coefficients are equal.

What would settle it

Simulate a known locally stationary Hawkes process with a non-constant Lipschitz reproduction rate $g^*$, compute the Remark 6 threshold $d \ge C_{\mu,\Phi} K^{-1} \inf_{C>0} \|g^*-C\|^4_{L^2[0,1]}$, fix $d$ above it, and record $\Lambda^d_T/T$ for growing $T$; Proposition 2 predicts a positive limit and a rejection probability tending to 1, so a flat likelihood ratio or bounded rejection rate across many long paths would refute the consistency claim.

Watch

Extended reading notes

Core claim

The paper's central claim is that the conventional Ogata MLE, applied to the locally stationary intensity $\lambda^T_{k,t}(\eta,g)=\Phi_k[\mu_k(t/T,\eta)+\sum_l \int_0^t g(t/T)\varphi_{kl}(t-s,\eta)\,dN^T_{l,s}]$ as if the process were stationary, is a legitimate estimator: Theorem 1 and Corollary 1 give $\sqrt{T}(\hat\vartheta_T-\vartheta^*)\to N(0,I(\vartheta^*)^{-1})$ for a single path when the true parameter is interior, with $I(\vartheta^*)$ the $x$-averaged stationary Fisher information. For the semi-nonparametric question, Theorem 2 shows the likelihood-ratio statistic $\Lambda^d_T$ for $g$ constant converges to $\chi^2(d)$ under the null, and Proposition 2 shows $\Lambda^d_T$ diverges under any non-constant continuous $g$, provided $d$ exceeds a threshold $d^*(\eta^*,g^*)$; Corollary 2 adjusts the degrees of freedom when nuisance interaction parameters sit on the boundary of the parameter space.

Load-bearing premise

The load-bearing premise is that the process stays subcritical at every rescaled time (each event triggers fewer than one descendant on average) and that, for the nonparametric test, the practitioner's fixed polynomial degree $d$ already exceeds an unknown threshold depending on the true reproduction rate---if $d$ is too small, the best degree-$d$ polynomial fit to the true rate can be constant and the likelihood-ratio gap collapses.

Editorial extensions

If this is right

  • A practitioner can reuse existing stationary-Hawkes MLE code on locally stationary data and still obtain $\sqrt{T}$ rates and Gaussian confidence intervals, without estimating local windows.
  • The likelihood-ratio test gives an asymptotically valid level-$\alpha$ test of constant reproduction rate, with $\chi^2(d)$ critical values, and a conservative boundary-corrected version when some interactions are zero.
  • Under the alternative, the test has power tending to 1 against every non-constant continuous reproduction rate, as long as the fixed polynomial degree is chosen above the threshold $d^*$; the required degree does not grow with $T$.
  • The reproduction-rate estimate $\hat g^d_T$ tracks smooth time variations of $g$ in simulations, suggesting the same polynomial machinery may support fully nonparametric estimation if the degree is allowed to grow with $T$.

Reading between the lines

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

  • The fixed-degree condition means a practitioner must choose $d$ before seeing the data, and the paper gives no data-driven way to know whether $d>d^*$; running the test over several degrees and checking that the MLE stays inside the parameter space is the practical safeguard its own simulations point to.
  • Because the threshold $d^*$ is governed by how far $g^*$ is from the constant functions (Remark 6 makes this explicit for Lipschitz $g$), the test is hardest precisely for slowly varying, nearly constant reproduction rates; power should increase with the $L^2$ distance from $g$ to the constants.
  • If the convergence of $\hat g^d_T$ could be extended to $d=d_T\to\infty$, the same likelihood-ratio construction would give a fully nonparametric test and an estimator of the integrated endogeneity $\int_0^1 g(x)\,dx$; the paper identifies this as future work.
  • For intraday power markets, rejection of constant $g$ in 20 of 21 sessions implies that time-dependent baseline models alone may miss an endogenous, participation-driven component of order flow; a direct follow-up would link the estimated $g(t/T)$ to time-to-delivery or market fundamentals.
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

4 major / 3 minor

Summary. The paper considers maximum likelihood estimation for a single observed path of a multivariate locally stationary Hawkes process with intensity λ^T_{k,t}(η,g)=Φ_k[μ_k(t/T,η)+Σ_l ∫_0^t g(t/T)φ_{kl}(t-s,η)dN^T_{l,s}]. In the fully parametric case, Theorem 1 and Corollary 1 claim √T-consistency and asymptotic normality of the Ogata MLE, extending earlier results of Ogata, Kwan, and Clinet-Yoshida. In the semi-nonparametric case, the authors propose a likelihood-ratio test based on Bernstein polynomials of fixed degree d to test whether the reproduction rate g is constant. Theorem 2 claims a χ²(d) null limit, and Proposition 2 claims that the test is consistent for every non-constant continuous g provided d exceeds a threshold d*(η*,g*). The paper also reports simulations and an application to German intraday power market order flow.

Significance. If the central results are correct, the paper makes a useful contribution by showing that the classical Ogata MLE remains √T-consistent and asymptotically normal in a locally stationary, multivariate, nonlinear Hawkes model, and by providing a simple misspecification-robust test for time dependence of the reproduction rate. The proof strategy follows the well-established Clinet-Yoshida framework with a locally stationary ergodic approximation, and the numerical/empirical sections give constructive evidence. However, the advertised nonparametric consistency claim is more delicate than the abstract suggests: it is not uniform over all continuous alternatives for a fixed implemented degree d, and the proof of Proposition 2 relies on assumptions that are not listed in its statement. These issues affect the paper's headline claims and need to be corrected before publication.

major comments (4)
  1. [Abstract and §1.2/Proposition 2] The abstract states that the likelihood-ratio test "remains consistent over the whole space of continuous functions of [0,1]". Proposition 2, however, only proves that for each fixed non-constant g* there exists d*(η*,g*) such that Λ_d^T diverges for every d>d*. The statistic Λ_d^T is computed with a degree d fixed before seeing the data, and no data-driven rule is given to ensure d exceeds d*. Thus for any fixed d there are continuous alternatives—for instance high-frequency, small-amplitude oscillations whose best degree-d Bernstein approximation is nearly constant—for which the likelihood-ratio gap in Lemma 13 collapses and the test has vanishing power. This is not a uniform consistency statement over C[0,1]. The abstract and Section 1.2 should be reworded to state exactly what is proved: for every non-constant continuous g*, there exists a sufficiently large degree d such that the test is consistent, and the practical limitation of the fixed-degree choice should be acknowledged, as the paper itself partly does in Remark 6.
  2. [Proposition 2 and Lemma 12] Proposition 2 is stated under Assumptions 1 to 3 only, but its proof invokes Assumption 11 (strict lower bound on Φ'_k) in Lemma 12 and uses Assumption 8 for the positivity of the denominator in the lower bound. Without Assumption 11, the key bound (28) may vanish and the strict positive gap K that drives Λ_d^T→∞ is not guaranteed. The statement of Proposition 2 should therefore include the assumptions actually used, for example Assumptions 6, 8, and 11, or the proof should be modified to avoid them. In addition, Assumption 11 is misprinted: it says "for any k=1...p" but the index k runs over the K coordinates of the process, not the dimension p of η.
  3. [§6.2, proof of Theorem 2] The proof of Theorem 2 contains inconsistent counting of the null restrictions. The null hypothesis in (11) is ϖ_0=⋯=ϖ_d, which imposes d restrictions on the d+1 Bernstein coefficients. In §6.2, however, the constrained set is written as Ξ_d^0={ϖ_0=⋯=ϖ_{d-1}}, and later the null constraints are stated as ϑ_{p+1}=⋯=ϑ_{p+d-1}=0, which is d−1 restrictions. The final degrees of freedom are then computed as k=p+d+1−(p+1)=d. These index shifts need to be aligned so that the chi-square limit has the correct number of degrees of freedom; as written, the proof does not rigorously establish χ²(d).
  4. [Corollary 2] Corollary 2, which accounts for nuisance parameters on the boundary of the parameter space, is not proved in the manuscript; it is deferred to the self-cited unpublished annex [Lotz, 2024]. Since the boundary issue is explicitly identified as important in Section 2.3 and is relevant to the empirical application, the paper should either include a proof of Corollary 2 or clearly label it as a result from an external unpublished manuscript. The statement also contains a typo: "for any sufficiently large n" should presumably read "for sufficiently large T", and the statistic is written as Λ_T rather than Λ_d^T.
minor comments (3)
  1. [§1.3] The sentence "These may be regarded as a standard approach within which falls our proof for Theorems and 1 and 2" contains a typo and should read "Theorems 1 and 2".
  2. [§3.1] The first sentence says "an empirical illustration of Theorem 2 and Proposition 3", but the relevant result for the simulation of the time-dependence test is Proposition 2, not the consistency result Proposition 3 in Section 6.1.
  3. [§2.3, Definition 1] The notation Ξ_d^0 is used in Definition 1 before it is formally defined; it would help to define it explicitly at the point of the null hypothesis (11).

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: the main CLT and consistency results are proved from stated assumptions; the only self-citation is an ancillary boundary correction, and the abstract's 'whole space' wording is an overstatement rather than a circularity.

full rationale

I find no circular step in the derivation chain. Theorem 1 and Corollary 1 are obtained by verifying the conditions of the Clinet-Yoshida M-estimator program through Proposition 1 (moment bounds), Proposition 2 (ergodic approximation), and Proposition 3 (consistency); each of these is proved using coupling, resolvent, and martingale arguments rather than assumed as an input. Proposition 2's test consistency is existential: for each non-constant continuous g*, the proof exhibits a degree threshold d* such that d > d* makes the likelihood ratio diverge. The gap is built from Lemmas 11-13 via KL-type lower bounds and Bernstein approximation, not from a fitted parameter. The abstract's claim that the test is 'consistent over the whole space of continuous functions' is stronger than the theorem, because any fixed-degree test is only guaranteed against alternatives whose true g* lies above the associated d*-threshold; this is a scope/overstatement issue, not circularity. The only self-citation is Corollary 2, whose boundary degrees-of-freedom correction is imported from [Lotz, 2024, Annex]; that result is ancillary, not used in the proofs of Theorem 1 or Proposition 2, so it does not make the paper's central claims self-referential. The omission of Assumption 11 from the statement of Proposition 2 is a rigor gap, not a circular step.

Assumptions & free parameters 2 free parameters · 8 assumptions · 0 invented entities

The theory relies on standard Hawkes stability, identifiability, and regularity conditions plus the Clinet-Yoshida M-estimator framework. The only user-chosen input in the testing procedure is the Bernstein degree d, and Proposition 2 requires it to exceed an unknown, g-dependent threshold. No new entities are introduced.

free parameters (2)
  • Bernstein degree d
    User-selected fixed degree for the reproduction-rate sieve; Proposition 2 requires d > d*(η*,g*), an unknown threshold depending on the true target.
  • Threshold d*(η*,g*)
    Exists in Proposition 2 but is not computable from data; Remark 6 gives only a Lipschitz-case bound, so it is an ad hoc parameter of the theorem.
assumptions (8)
  • domain assumption Assumption 1: Φ_k are 1-Lipschitz and non-negative; non-inhibitive interactions.
    Positivity and Lipschitz continuity are needed for the coupling and renewal arguments; inhibitory interactions are excluded, as admitted in Section 5.1.
  • domain assumption Assumption 2: sup_{x∈[0,1]} g*(x) ρ(∫ φ_kl(s,η*) ds) < 1.
    Subcriticality at all normalized times guarantees existence of stationary embedded comparison processes and uniform moment bounds; all lemmas fail without it.
  • domain assumption Assumption 3: no φ(·,η1) is proportional to φ(·,η2) when η1≠η2.
    Identifiability of the reproduction rate scale; without it g and φ can trade off and the test target is not well defined.
  • domain assumption Assumption 8: inf_{η,x} Φ_k[μ_k(x,η)] > 0.
    Keeps intensity bounded away from zero, used in the score, Lyapunov and Lindeberg bounds.
  • domain assumption Assumption 10: the local Fisher information I^(k)(x,ϑ) is positive definite for some x,k.
    Non-degenerate information matrix is required for the inversion step in the CLT.
  • domain assumption Assumption 11: Φ'_k(x) > ε > 0.
    Strict convexity and bijectivity are used in Lemma 12 to obtain a positive likelihood gap between a non-constant g and the best constant rate; omitted from the abstract.
  • standard math Theorem 7 of Brémaud and Massoulié (1996): stationary nonlinear Hawkes processes exist on the same Poisson base.
    Load-bearing for constructing the stationary embedded processes in (4) and for ergodicity in Lemma 3.
  • standard math Clinet-Yoshida (2017) Theorem 3.11 and Lemma 3.10: M-estimator asymptotic framework.
    The MLE CLT is obtained by checking their conditions rather than reproving the master theorem.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A non-local estimator for locally stationary Hawkes processes." pith.science (2026). https://pith.science/paper/5HFUGJ4C

@misc{pith2026250602631,
  author       = {Pith},
  title        = {Pith review of: A non-local estimator for locally stationary Hawkes processes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5HFUGJ4C}},
  note         = {Machine review of arXiv:2506.02631}
}
abstract

We consider the problem of estimating the parameters of a non-stationary Hawkes process with time-dependent reproduction rate and baseline intensity. Our approach relies on the standard maximum likelihood estimator (MLE), coinciding with the conventional approach for stationary point processes characterised by [Ogata, 1978]. In the fully parametric setting, we find that the MLE over a single observation of the process over $[0, T]$ remains consistent and asymptotically normal as $T \to \infty$. Our results extend partially to the semi-nonparametric setting where no specific shape is assumed for the reproduction rate $g \colon [0, 1] \mapsto \mathbb{R}_+$. We construct a time invariance test with null hypothesis that g is constant against the alternative that it is not, and find that it remains consistent over the whole space of continuous functions of [0, 1]. As an application, we employ our procedure in the context of the German intraday power market, where we provide evidence of fluctuations in the endogeneity rate of the order flow.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

34 extracted references · 33 canonical work pages

  1. [1]

    Adams, R. A. and Fournier, J. F. (2003). Sobolev spaces . Elsevier Science, second edition

    work page 2003

  2. [2]

    and Blanc, P

    Alfonsi, A. and Blanc, P. (2015). Dynamic optimal execution in a mixed-market-impact hawkes price model

    work page 2015

  3. [3]

    Andrews, D. (1999). Estimation when a parameter is on a boundary. Econometrica , 67(6)

    work page 1999

  4. [4]

    Bacry, E., Delattre, S., Hoffmann, M., and Muzy, F. (2013). Some limit theorems for H awkes processes and application to financial statistics. Stochastic Processes and their Applications , 123(7):2475--2499

    work page 2013

  5. [5]

    Bacry, E., Delattre, S., Hoffmann, M., and Muzy, J. F. (2011). Modeling microstructure noise with mutually exciting point processes

    work page 2011

  6. [6]

    Bonnet, A., Martinez Herrera, M., and Sangnier, M. (2023). Inference of multivariate exponential H awkes processes with inhibition and application to neuronal activity. Statistics and Computing , 33(4)

    work page 2023

  7. [7]

    and Massouli \'e , L

    Br \'e maud, P. and Massouli \'e , L. (1996). Stability of nonlinear hawkes processes. The Annals of Probability , pages 1563--1588

    work page 1996

  8. [8]

    Brignone, R., Gonzato, L., and Sgarra, C. (2024). Hawkes Processes in Energy Markets: Modelling, Estimation and Derivatives Pricing . Springer

    work page 2024

Show all 34 references
  1. [9]

    Brémaud, P. (2020). Point Process Calculus in Time and Space , volume 98 of Probability Theory and Stochastic Modelling . Springer

    2020

  2. [10]

    and Hall, P

    Chen, F. and Hall, P. (2013). Inference for a nonstationary self-exciting point process with an application in ultra-high frequency financial data modeling . Journal of Applied Probability , 50(4):1006--1024

    2013

  3. [11]

    Chevalier, E., Hafsi, Y., and Vath, V. L. (2023). Uncovering market disorder and liquidity trends detection

    2023

  4. [12]

    and Potiron, Y

    Clinet, S. and Potiron, Y. (2018). Estimating the integrated parameter of the time-varying parameter self-exciting process. arXiv: Statistical Finance

    2018

  5. [13]

    and Yoshida, N

    Clinet, S. and Yoshida, N. (2017). Statistical inference for ergodic point processes and application to limit order book. Stochastic Processes and their Applications , 127(6):1800--1839

    2017

  6. [14]

    An Introduction to the Theory of Point Processes: elementary theory and methods

    Daley and Vere-Jones (2002). An Introduction to the Theory of Point Processes: elementary theory and methods . Probability and its Applications. Springer, second edition

    2002

  7. [15]

    and Gruet, P

    Deschatre, T. and Gruet, P. (2023). Electricity intraday price modelling with marked H awkes processes. Applied Mathematical Finance , pages 227--260

    2023

  8. [16]

    Deschatre, T., Gruet, P., and Lotz, A. (2025). Some limit theorems for locally stationary hawkes processes

    2025

  9. [17]

    Geyer, C. (1994). On the Asymptotics of Constrained M-Estimation . The Annals of Statistics , 22(4):1993--2010

    1994

  10. [18]

    O., and Staffans, O

    Gripenberg, G., Londen, S. O., and Staffans, O. (1990). Volterra Integral and Functional Equations . Encyclopedia of Mathematics and its Applications. Cambridge University Press

    1990

  11. [19]

    and Oakes, D

    Hawkes, A. and Oakes, D. (1974). A cluster process representation of a self-exciting process. Journal of Applied Probability , 11(3):493--503

    1974

  12. [20]

    and Shiryaev, A

    Jacod, J. and Shiryaev, A. (1987). Limit theorems for stochastic processes . Springer-Verlag

    1987

  13. [21]

    and Rosenbaum, M

    Jaisson, T. and Rosenbaum, M. (2015). Limit theorems for nearly unstable H awkes processes. The Annals of Applied Probability , 25(2)

    2015

  14. [22]

    and Kiesel, R

    Kramer, A. and Kiesel, R. (2021). Exogenous factors for order arrivals on the intraday electricity market. Energy Economics , 97

    2021

  15. [23]

    Kwan, J. (2023). Asymptotic analysis and ergodicity of the Hawkes process and its extensions . PhD thesis, UNSW Sydney

    2023

  16. [24]

    and Schilling, R

    Kühn, F. and Schilling, R. L. (2023). Maximal inequalities and some applications

    2023

  17. [25]

    A sparsity test for hawkes processes

    Lotz (2024). A sparsity test for hawkes processes. arXiv

    2024

  18. [26]

    Mammen, E. (2017). Nonparametric estimation of locally stationary hawkes processe

    2017

  19. [27]

    Ogata, Y. (1978). The asymptotic behaviour of maximum likelihood estimators for stationary point processes. Ann Inst Stat Math , 30:243--261

    1978

  20. [28]

    Ogata, Y. (1988). Statistical models for earthquake occurrences and residual analysis for point processes. Journal of the American Statistical Association , 83(401):9--27

    1988

  21. [29]

    Reynaud-Bouret, P., Rivoirard, V., and Tuleau-Malot, C. (2013). Inference of functional connectivity in neurosciences via H awkes processes. In 2013 IEEE Global Conference on Signal and Information Processing

    2013

  22. [30]

    Rio, E. (2017). Asymptotic Theory of Weakly Dependent Random Processes. Springer

    2017

  23. [31]

    and Sachs, R

    Roueff, F. and Sachs, R. V. (2019). Time-frequency analysis of locally stationary hawkes processes. Bernoulli , 23(2):1355--1385

    2019

  24. [32]

    Roueff, F., von Sachs , R., and Sansonnet, L. (2016). Locally stationary hawkes processes. Stochastic Processes and their Applications , 126(6):1710--1743

    2016

  25. [33]

    Toke, I. M. (2011). ``Market Making'' in an Order Book Model and Its Impact on the Spread , pages 49--64. Springer Milan, Milano

    2011

  26. [34]

    Van den Vaart, A. (1998). Asymptotic statistics . Cambrige Series in statistical and probabilistic mathematics

    1998

Pith tools

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