Maxima of stationary systems of randomly time-changed L\'evy particles
Pith reviewed 2026-05-10 15:22 UTC · model grok-4.3
The pith
Systems of randomly time-changed Lévy particles generate stationary max-infinitely divisible processes whose extremes are attracted to Lévy-Brown-Resnick processes.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
We construct stationary max-infinitely divisible processes from systems of randomly time-changed Lévy particles. Classical examples without time change are max-stable up to marginal transformations. Random time change alters the dependence structure leading beyond the max-stable setting while preserving stationarity through reconfiguration of starting points. The extremal behavior links to an associated max-stable Lévy-Brown-Resnick process through the max-domain of attraction, yielding new non-trivial stationary processes in that domain.
What carries the argument
A system of randomly time-changed Lévy particles with reconfigured starting points to preserve stationarity, which generates a max-id process in the MDA of a Lévy-Brown-Resnick process.
If this is right
- The dependence structure of the max-id process differs from that of max-stable processes due to the random time changes.
- The process remains stationary despite the random time changes.
- The finite-dimensional distributions belong to the max-domain of attraction of the associated Lévy-Brown-Resnick process.
- This method produces a large class of stationary processes previously scarce in the literature for the MDA.
Where Pith is reading between the lines
- Extending random time changes to other Lévy processes could produce processes with varied dependence in their MDA.
- The particle system approach may aid in simulating non-max-stable stationary extremes for applications in environmental modeling.
- Connecting this to potential theory could reveal new ways to compute extremal indices for time-changed systems.
Load-bearing premise
A suitable reconfiguration of the starting points of the particle system preserves stationarity for the chosen random time changes of the underlying Lévy particles.
What would settle it
Simulate the maxima of the time-changed particle system for large thresholds and check whether their normalized distributions converge to those of the associated Lévy-Brown-Resnick process; failure to converge or loss of stationarity without the reconfiguration would disprove the claims.
read the original abstract
In this work, we consider maxima of systems of randomly time-changed L\'evy particles. We give a general construction to obtain infinite-dimensional classes $\{Z^{\alpha}\}$ of stationary max-infinitely divisible (max-id) processes. These classes are indexed by admissible mass functions $\alpha$, which induce state-dependent time changes of the underlying L\'evy particles. This gives a generalization of the well-known (L\'evy--)Brown--Resnick process $Z^{1}$. In contrast to $\alpha\equiv 1$, the variability of non-constant mass functions $\alpha$ changes the dependence structure of the max-id process and goes beyond the max-stable setting while preserving stationarity. We then explore the extent of the so-called max-domain of attraction (MDA) of a given (L\'evy--)Brown--Resnick process $Z^1$, by studying convergence of rescaled maxima of independent copies of $Z^{\alpha}$ to $Z^1$. Thus, our work combines potential theory for Markov processes and extreme value theory to yield a novel, infinite-dimensional, and interpretable class $\{Z^{\alpha}\}$ of stationary processes in the MDA of a given (L\'evy--)Brown--Resnick process $Z^{1}$. So far, results on the extent of such domains have been scarce in the literature.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript constructs stationary max-infinitely divisible (max-id) processes from systems of randomly time-changed Lévy particles. Random time changes are shown to alter the dependence structure beyond the max-stable class, while stationarity is restored by reconfiguring the initial positions of the particles. The resulting processes are proved to lie in the max-domain of attraction (MDA) of an associated Lévy-Brown-Resnick process, via a combination of potential theory for the underlying Markov processes and extreme-value theory controlling finite-dimensional distributions and extremal dependence.
Significance. If the central claims hold, the work supplies a flexible and large class of new, non-trivial stationary max-id processes in the MDA of a given Lévy-Brown-Resnick process. This is a meaningful advance because concrete examples of processes in the MDA remain scarce; the explicit use of potential theory to handle the time-changed Markovian particles while preserving the MDA link is a clear technical strength.
major comments (2)
- [§2] The central stationarity claim rests on the reconfiguration of starting points (abstract and §2). The proof that this reconfiguration restores stationarity for arbitrary random time changes must explicitly verify that the finite-dimensional distributions remain invariant; without a quantitative bound on the discrepancy introduced by the time change, the claim that stationarity is preserved for the full class is not yet load-bearing.
- [§4] The MDA result (abstract and §4) asserts that the time-changed process belongs to the MDA of the Lévy-Brown-Resnick process. The argument must include an explicit control on the normalizing sequences and the convergence of the point processes; the current sketch does not rule out that the random time change could push the process outside the MDA for some choices of the Lévy measure.
minor comments (2)
- [Introduction] The notation for the random time-change process and the reconfigured initial measure should be introduced with a single consistent symbol set in the introduction to avoid later ambiguity.
- A short table comparing the dependence structure (e.g., extremal coefficient or pairwise tail dependence) of the time-changed process versus the original Brown-Resnick process would improve readability.
Simulated Author's Rebuttal
We thank the referee for the careful reading of our manuscript and the constructive comments. We address each major comment below, indicating the revisions we will undertake to strengthen the presentation.
read point-by-point responses
-
Referee: [§2] The central stationarity claim rests on the reconfiguration of starting points (abstract and §2). The proof that this reconfiguration restores stationarity for arbitrary random time changes must explicitly verify that the finite-dimensional distributions remain invariant; without a quantitative bound on the discrepancy introduced by the time change, the claim that stationarity is preserved for the full class is not yet load-bearing.
Authors: We appreciate the referee's emphasis on making the stationarity argument fully rigorous. The reconfiguration of initial positions is chosen precisely so that the Markovian structure of the underlying Lévy particles, combined with the potential-theoretic representation of the time-changed paths, yields shift-invariant finite-dimensional distributions. While the current proof invokes these properties to conclude invariance, we agree that an expanded, self-contained verification of the f.d.d. together with explicit quantitative bounds on the discrepancy induced by the random time change would render the argument more transparent. In the revised manuscript we will insert a dedicated lemma in §2 that (i) computes the finite-dimensional distributions after reconfiguration, (ii) shows their invariance under arbitrary time shifts, and (iii) supplies a uniform bound on the total-variation distance between the original and time-changed marginals, derived from the moment assumptions on the time-change process. revision: yes
-
Referee: [§4] The MDA result (abstract and §4) asserts that the time-changed process belongs to the MDA of the Lévy-Brown-Resnick process. The argument must include an explicit control on the normalizing sequences and the convergence of the point processes; the current sketch does not rule out that the random time change could push the process outside the MDA for some choices of the Lévy measure.
Authors: We concur that the MDA proof requires additional explicit controls. The existing argument uses potential theory to identify the extremal index and the limiting Poisson point process, but the normalizing sequences and the convergence in the space of point measures are only sketched. In the revision we will (i) state the precise normalizing sequences (which remain the same as those of the underlying Lévy-Brown-Resnick process because the time change has finite moments), (ii) prove convergence of the rescaled point processes by combining the continuous-mapping theorem with the already-established convergence of the time-changed Markov processes in the Skorokhod space, and (iii) add a short argument showing that, for every Lévy measure satisfying the integrability conditions of the paper, the random time change cannot eject the process from the MDA; the key estimate is a uniform bound on the tail of the time-changed exceedance measure that follows from the same potential-theoretic comparison used for stationarity. revision: yes
Circularity Check
No significant circularity; derivation self-contained
full rationale
The paper constructs stationary max-id processes from randomly time-changed Lévy particles, restores stationarity via reconfiguration of initial positions, and establishes MDA membership to a Lévy-Brown-Resnick process by combining potential theory for the underlying Markov processes with extreme-value theory for finite-dimensional distributions and extremal dependence. All steps are presented as explicit constructions and proofs without any reduction of a claimed prediction or uniqueness result to a fitted parameter, self-defined quantity, or load-bearing self-citation chain; the central claims rest on independent application of standard external tools rather than internal re-labeling or forced equivalence.
Axiom & Free-Parameter Ledger
axioms (2)
- standard math Standard properties of Lévy processes and Markov processes
- standard math Theory of max-infinitely divisible and max-stable processes
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.