arxiv: 2605.07059 · v1 · submitted 2026-05-08 · 🧮 math.PR

Recognition: 2 theorem links

· Lean Theorem

Modified ruin probability for a Cram\'er-Lundberg model driven by a compound mixed Poisson process

Momoka Tashiro, Noriyoshi Sakuma

Pith reviewed 2026-05-11 02:20 UTC · model grok-4.3

classification 🧮 math.PR

keywords ruin probabilityCramér-Lundberg modelmixed Poisson processsubexponential distributionheavy-tailed asymptoticslight-tailed regimeasymptotic equivalence

0 comments

The pith

In a Cramér-Lundberg model with mixed Poisson claim arrivals, the modified ruin probability is asymptotically equivalent to the classical one when claims are heavy-tailed, provided the mixing distribution stays below the net-profit boundary

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper studies ruin probabilities in an insurance risk model where claims arrive according to a compound mixed Poisson process, so the intensity is drawn from a mixing distribution rather than fixed. It proves that in the heavy-tailed regime, when the integrated claim-size distribution is subexponential and the upper endpoint of the mixing distribution lies below the net-profit boundary, the modified ruin probability and the classical ruin probability are asymptotically equivalent. In the light-tailed regime the authors establish a fixed-intensity ratio theorem together with an endpoint-atom result and a sharp endpoint-density asymptotic carrying an explicit constant.

Core claim

In the heavy-tailed regime, if the integrated claim-size distribution is subexponential and the upper endpoint of the mixing distribution stays below the net-profit boundary, the modified and classical ruin probabilities are asymptotically equivalent in the Cramér-Lundberg model driven by a compound mixed Poisson process. In the light-tailed regime, a fixed-intensity ratio theorem holds along with endpoint-atom and sharp endpoint-density results with explicit constants.

What carries the argument

The compound mixed Poisson process, which replaces the fixed intensity of a classical Poisson process with a random intensity drawn from a mixing distribution.

If this is right

Ruin calculations in heavy-tailed mixed-Poisson models reduce to standard subexponential formulas without further adjustment.
Light-tailed analysis yields explicit ratio constants and boundary asymptotics for both atoms and densities.
The results separate the effect of intensity mixing from the claim-size tail behavior.
Endpoint conditions on the mixing distribution become decisive for the equivalence.

Where Pith is reading between the lines

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

The equivalence implies that standard heavy-tailed ruin approximations remain usable even when arrival intensity is uncertain, provided the mixing range is restricted.
Similar mixing arguments could be applied to other risk processes such as those with stochastic premiums or interest rates.
Numerical verification of the boundary condition in insurance data sets would test whether the asymptotic equivalence appears in practice.

Load-bearing premise

The upper endpoint of the mixing distribution must remain below the net-profit boundary and the integrated claim-size distribution must be subexponential.

What would settle it

A simulation or explicit calculation in which the upper endpoint of the mixing distribution exceeds the net-profit boundary, showing that the modified ruin probability ceases to be asymptotically equivalent to the classical one.

read the original abstract

We study modified ruin probabilities in a Cram\'er-Lundberg model driven by a compound mixed Poisson process. In the heavy-tailed regime, if the integrated claim-size distribution is subexponential and the upper endpoint of the mixing distribution stays below the net-profit boundary, the modified and classical ruin probabilities are asymptotically equivalent. In the light-tailed regime, we prove a fixed-intensity ratio theorem and obtain both an endpoint-atom result and a sharp endpoint-density asymptotic with an explicit constant.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper gives explicit asymptotic equivalences for modified ruin probabilities in a mixed-Poisson Cramér-Lundberg model, with the heavy-tailed case reducing cleanly to the classical one under stated conditions.

read the letter

The central result is that when claim sizes have a subexponential integrated distribution and the mixing distribution on the intensity has upper endpoint strictly below the net-profit boundary, the modified ruin probability is asymptotically equivalent to the ordinary one. In the light-tailed regime the paper adds a fixed-intensity ratio theorem plus an endpoint-atom result and a sharp density asymptotic carrying an explicit constant. These statements are new in the mixed-Poisson setting and are presented with the precise hypotheses needed to keep the process subcritical. The conditions are not hidden; they are exactly what prevent the random intensity from driving the model into a different regime, so the reduction to classical ruin theory is transparent rather than forced. The explicit constant in the light-tailed density asymptotic is a concrete addition that earlier fixed-intensity work did not always supply. The logical structure looks sound from the stated theorems, with no circularity or unstated uniformity arguments required. The main limitation is that the endpoint restriction on the mixing distribution keeps the model from exploring truly supercritical random-intensity behavior, so the contribution stays within the usual subexponential and Cramér-Lundberg toolkit rather than breaking new ground on variable drift. This is useful work for specialists in insurance mathematics and applied probability who already follow ruin asymptotics. It is not broad enough to interest a general probability audience, but the claims are precise enough that a serious referee could check the derivations and the constants without undue effort. I would send it to peer review.

Referee Report

0 major / 3 minor

Summary. The manuscript examines modified ruin probabilities in a Cramér-Lundberg model driven by a compound mixed Poisson process. In the heavy-tailed regime, under the conditions that the integrated claim-size distribution is subexponential and the upper endpoint of the mixing distribution lies strictly below the net-profit boundary, the modified and classical ruin probabilities are shown to be asymptotically equivalent. In the light-tailed regime, the paper establishes a fixed-intensity ratio theorem, an endpoint-atom result, and a sharp asymptotic for the endpoint density with an explicit constant.

Significance. If the derivations hold, the work extends classical ruin asymptotics to mixed-Poisson intensity models with explicit, verifiable conditions and constants. The heavy-tailed equivalence reduces to the standard compound-Poisson case precisely when the mixing cannot drive the process supercritical, while the light-tailed results supply concrete ratios and expansions useful for risk calculations. The explicit hypotheses and constants constitute a clear technical contribution to insurance mathematics.

minor comments (3)

§2, Definition 2.3: the notation for the modified ruin probability ψ_θ(u) is introduced without an immediate comparison to the classical ψ(u), which would help readers track the distinction throughout the proofs.
Theorem 3.2 (light-tailed fixed-intensity ratio): the statement of the ratio limit is clear, but the proof sketch in the text does not explicitly indicate where the dominated-convergence step is justified under the given moment assumptions.
Figure 1: the caption does not state the parameter values used for the numerical illustration of the heavy-tailed equivalence, making it difficult to reproduce or compare with the asymptotic formula.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the careful reading and positive evaluation of our work, including the accurate summary of the heavy-tailed asymptotic equivalence and the light-tailed results (fixed-intensity ratio, endpoint-atom, and sharp density asymptotic). The recommendation for minor revision is noted, but no specific major comments were provided in the report.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper's central claims consist of asymptotic equivalences and explicit tail asymptotics for ruin probabilities, conditioned on standard external assumptions (subexponentiality of the integrated claim-size distribution and the mixing distribution's upper endpoint lying strictly below the net-profit boundary). These hypotheses are stated explicitly in the abstract and theorems; they are not derived from the paper's own equations or fitted parameters. Light-tailed results are likewise given with explicit constants and ratios. No load-bearing step reduces by construction to a self-definition, a fitted input renamed as a prediction, or a self-citation chain. The derivation chain is therefore self-contained against external benchmarks in probability theory.

Axiom & Free-Parameter Ledger

0 free parameters · 3 axioms · 0 invented entities

The paper rests on standard mathematical properties of subexponential distributions and the classical net-profit condition of the Cramér-Lundberg model. No free parameters are fitted, no new entities are postulated, and no ad-hoc axioms beyond domain-standard assumptions appear in the abstract.

axioms (3)

domain assumption Integrated claim-size distribution is subexponential
Invoked to obtain asymptotic equivalence in the heavy-tailed regime.
domain assumption Upper endpoint of mixing distribution lies below net-profit boundary
Required for the equivalence statement; appears as a load-bearing premise.
standard math Standard properties of compound Poisson processes and ruin probabilities
Background framework assumed throughout.

pith-pipeline@v0.9.0 · 5371 in / 1467 out tokens · 40434 ms · 2026-05-11T02:20:49.187373+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear
Theorem 3.1. Assume that ℓ_1 < c/μ, that F_I ∈ S, and that lim_{y→−∞} ψ(y)=1. Then ψ(u)∼ψ_cl(u)∼E[Λμ/(c−Λμ)] F_I(u)
IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear
Theorem 4.3 ... ψ(u)∼C_ℓ1 B C_1 Γ(b) / (D_1 u)^b e^{-R_1 u}

Reference graph

Works this paper leans on

8 extracted references · 8 canonical work pages

[1]

Aurzada and M

F. Aurzada and M. Buck, Ruin probabilities in the Cram´ er–Lundberg model with temporarily negative capital.Eur. Actuar. J.10(2020), no. 1, 261–269

work page 2020
[2]

Dassios and S

A. Dassios and S. Wu, Parisian ruin with exponential claims, LSE Research Online Documents on Economics, no. 32033 (2008), available at https://eprints.lse.ac.uk/ 32033/1/Parisian_ruin_with_exponential_claims_(LSERO).pdf

work page 2008
[3]

Gu´ erin and J.-F

H. Gu´ erin and J.-F. Renaud, On the distribution of cumulative Parisian ruin.Insurance Math. Econom.73(2017), 116–123

work page 2017
[4]

Grandell,Mixed Poisson Processes

J. Grandell,Mixed Poisson Processes. Monographs on Statistics and Applied Probability,

work page
[5]

Chapman & Hall, London, 1997

work page 1997
[6]

Mikosch,Non-life Insurance Mathematics: An Introduction with the Poisson Process

T. Mikosch,Non-life Insurance Mathematics: An Introduction with the Poisson Process. 2nd ed., Universitext, Springer, Berlin, 2009

work page 2009
[7]

Asmussen and H

S. Asmussen and H. Albrecher,Ruin Probabilities. 2nd ed., Advanced Series on Statistical Science & Applied Probability, 14. World Scientific, Singapore, 2010

work page 2010
[8]

Schmidli, On the distribution of the surplus prior and at ruin.ASTIN Bull.29(1999), 227–244

H. Schmidli, On the distribution of the surplus prior and at ruin.ASTIN Bull.29(1999), 227–244. 6

work page 1999