Local tail asymptotics for the joint distribution of length and of maximum of a random walk excursion
Pith reviewed 2026-05-25 08:46 UTC · model grok-4.3
The pith
The joint distribution of length, maximum, and time of maximum for a random walk excursion admits explicit local tail asymptotics under negative drift.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
We determine the local asymptotics of the joint distribution of the length, maximum and the time at which this maximum is achieved. This result allows one to obtain a local central limit theorems for the length of the excursion conditioned on large values of the maximum.
What carries the argument
The local asymptotic formula for the joint probabilities of excursion length, height, and the step at which the height is attained.
If this is right
- A local central limit theorem holds for the length of the excursion conditioned on the maximum being large.
- The conditional law of the time of the maximum given length and height is described asymptotically.
- The same asymptotic machinery yields tail probabilities for the joint event that length exceeds n and maximum exceeds x.
Where Pith is reading between the lines
- The same local asymptotics should hold, after suitable scaling, for the continuous-time analogue of Brownian motion with negative drift.
- The formulas could be used to design efficient importance-sampling schemes that generate excursions with atypically large maxima.
- Extensions to other additive functionals of the excursion path, such as its area, appear feasible with the same conditioning technique.
Load-bearing premise
The random walk has negative drift and light-tailed increments.
What would settle it
Numerical counts of excursion length and maximum in long simulated paths of a negative-drift walk with light tails that fail to match the predicted asymptotic joint frequencies within the stated error term.
read the original abstract
This note is devoted to the study of the maximum of the excursion of a random walk with negative drift and light-tailed increments. More precisely, we determine the local asymptotics of the joint distribution of the length, maximum and the time at which this maximum is achieved. This result allows one to obtain a local central limit theorems for the length of the excursion conditioned on large values of the maximum.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper derives local tail asymptotics for the joint distribution of the length, maximum value, and argmax time of an excursion of a random walk with negative drift and light-tailed increments. These asymptotics are then applied to obtain local central limit theorems for the conditional distribution of the excursion length given that the maximum is large.
Significance. If the derivations hold, the results supply precise local asymptotics in a standard regime for random walk excursions, extending beyond global tail estimates to joint laws and conditional local CLTs. This level of detail is useful for applications requiring conditioned path properties under exponential moments.
minor comments (2)
- The abstract states that the joint asymptotics 'allow one to obtain' the conditional local CLTs, but the precise mechanism (e.g., which marginal or joint term is inverted) is not previewed; a one-sentence outline in the abstract or introduction would improve readability.
- Notation for the excursion length, maximum, and argmax (likely denoted something like (τ, M, θ) or similar) should be introduced explicitly at the first use in §1 to avoid any ambiguity when the joint local limit is stated.
Simulated Author's Rebuttal
We thank the referee for the positive assessment of the manuscript and the recommendation to accept.
Circularity Check
No significant circularity detected
full rationale
The paper derives local tail asymptotics for the joint law of (length, maximum, argmax) of excursions for random walks with negative drift and light-tailed increments, plus a conditional local CLT corollary. The abstract and described argument structure rest on standard random-walk tail and renewal theory under the stated assumptions; no equation or step is shown to reduce by construction to a fitted parameter, self-definition, or unverified self-citation chain. The derivation is therefore self-contained against external benchmarks.
Axiom & Free-Parameter Ledger
axioms (1)
- domain assumption Random walk has negative drift and light-tailed increments
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.