Branching Process in a Varying Environment: How to Grow Like the Product of Means

A. Shklyaev; Y. Kirpicheva

arxiv: 2604.05999 · v1 · submitted 2026-04-07 · 🧮 math.PR

Branching Process in a Varying Environment: How to Grow Like the Product of Means

Y. Kirpicheva , A. Shklyaev This is my paper

Pith reviewed 2026-05-10 19:08 UTC · model grok-4.3

classification 🧮 math.PR

keywords branching processvarying environmentrandom environmentmartingale limitsurvival probabilitysufficient conditionsoffspring distributionproduct of means

0 comments

The pith

New sufficient conditions ensure that in a branching process with varying environment the martingale limit W is positive exactly on the survival event, without requiring finite second moments for every generation.

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

The paper develops new sufficient conditions under which a branching process in a varying environment satisfies that the normalized population size Z_n divided by its expectation converges to a random variable W that is strictly positive with probability exactly equal to the probability of survival. This implies that, conditional on survival, the population grows asymptotically like the product of the successive mean offspring numbers. The conditions are designed to work for random environments and, crucially, do not require the second moments of the offspring distributions to be finite in every generation. A reader cares because many natural models of population growth allow heavy-tailed offspring numbers, and the new conditions remove an artificial barrier that had restricted earlier results.

Core claim

We find a new kind of sufficient conditions, applicable to branching processes in a random environment, such that P(W > 0) equals the survival probability, where W is the almost-sure limit of the natural martingale Z_n / E Z_n; an important property of these estimates is that they do not necessarily assume that E X_{i,1}^2 are finite for every i.

What carries the argument

The new sufficient conditions on the sequence of offspring distributions and the environment that guarantee the martingale limit W is positive precisely on the event of non-extinction.

If this is right

Conditional on survival the population size Z_n is asymptotically equivalent to W times the product of the mean offspring numbers.
The result applies directly to branching processes whose environment is itself random.
The same conclusion holds even when the second moment of the offspring distribution is infinite for some generations.
The martingale convergence alone, under these conditions, implies that the process cannot survive at a smaller order than its expectation.

Where Pith is reading between the lines

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

The relaxation of the second-moment requirement opens the door to modeling populations that occasionally produce very large broods while still guaranteeing growth proportional to the mean.
Similar moment-free conditions might be transferable to other multiplicative processes such as random walks in random media or certain multiplicative cascades.
Simulation experiments with Pareto-tailed offspring distributions satisfying the new conditions could provide a practical check on the boundary between survival and extinction.

Load-bearing premise

The new sufficient conditions hold for the given sequence of offspring distributions and the environment.

What would settle it

An explicit sequence of offspring distributions and environment for which the new conditions are satisfied yet P(W>0) is strictly smaller than the survival probability.

read the original abstract

Consider a branching process $\{Z_n\}$ in a varying environment. Let $\{W_n\}$ be the natural martingale $Z_n/{\bf E}Z_n$. It converges to some random variable $W$ as $n\to\infty$. An important problem is to show that ${\bf P}(W>0)$ equals the survival probability, so that $Z_n$ is either $0$ or of the order ${\bf E}Z_n$. We find a new kind of sufficient conditions, applicable to branching processes in a random environment. An important property of our estimates is that we don't necessary assume that ${\bf E}X_{i,1}^2$ are finite for every $i$.

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 sufficient conditions for non-degenerate martingale limits in branching processes in varying or random environments without needing finite second moments.

read the letter

The main point is that the authors supply a new set of sufficient conditions under which the normalized branching process Z_n / E Z_n converges to a limit W that is positive exactly on the survival set. These conditions apply to random environments and replace the usual finite-variance requirement with weaker integrability controls on truncated offspring distributions, while still using the product of the means as the growth benchmark. The construction starts from the standard martingale convergence, which holds regardless of moments, and then adds tail-type bounds to get non-degeneracy. That step looks clean and avoids circularity. The estimates are stated explicitly in the main theorem, which is a plus for anyone who wants to check them directly. The approach is technically sound on its own terms and does relax a common restriction in the literature. A minor limitation is that the summary gives no concrete examples where the new conditions hold but second moments fail, so it is still unclear how much wider the applicability becomes in practice. If the full paper includes such checks or shows the conditions are easy to verify for common distributions, that would make the gain clearer. This is a focused technical note for people working on limit theorems for branching processes in random media. A specialist reader would get direct value from the precise statement of the conditions and the proof strategy. I would send it to peer review; the core claim is grounded enough to deserve referee time on the details of the estimates.

Referee Report

0 major / 2 minor

Summary. The manuscript studies branching processes {Z_n} in a varying environment (including random environments). It considers the natural martingale W_n = Z_n / E[Z_n] which converges almost surely to a limit W, and derives new sufficient conditions ensuring that P(W > 0) equals the survival probability of the process. These conditions are formulated in terms of integrability properties of the normalized offspring distributions and are designed to apply without requiring E[X_{i,1}^2] < ∞ for each generation i.

Significance. The result relaxes a common second-moment assumption in the literature on branching processes in random environments by replacing variance bounds with weaker tail controls on the offspring distributions. This could enable analysis of models with heavy-tailed reproduction laws. The explicit statement of the integrability conditions in the main theorem is a strength, as is the separation of the always-valid martingale convergence from the non-degeneracy argument.

minor comments (2)

Abstract: The abstract states the main claim but provides neither a sketch of the argument nor the precise form of the new integrability conditions. Adding one or two sentences summarizing the key hypotheses would improve accessibility.
The manuscript does not include concrete examples or numerical checks verifying that the new conditions can be satisfied by non-trivial offspring distributions with infinite variance. While not required for the theorem, such illustrations would help readers gauge restrictiveness.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive and accurate summary of our manuscript, which correctly identifies the main contribution: new sufficient conditions for non-degeneracy of the martingale limit W in branching processes in varying (including random) environments, without requiring finite second moments. We are pleased that the relaxation of the common E[X_{i,1}^2] < ∞ assumption and the explicit integrability conditions are viewed as strengths.

Circularity Check

0 steps flagged

No significant circularity; derivation self-contained

full rationale

The paper's central result introduces explicit integrability conditions on normalized offspring distributions to ensure non-degeneracy of the martingale limit W without requiring finite second moments. Martingale convergence itself is invoked as a standard fact independent of the new conditions, and the sufficient conditions are stated directly in the main theorem using truncated moments and product-of-means growth. No step reduces the claimed result to a fit, self-definition, or load-bearing self-citation chain; the argument remains externally verifiable against the stated assumptions.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The result rests on standard martingale convergence for non-negative supermartingales and on the definition of the branching process in a varying environment. No free parameters or invented entities are introduced in the abstract.

axioms (2)

standard math The normalized process W_n is a non-negative martingale and therefore converges almost surely to a limit W.
Invoked implicitly when discussing convergence of W_n to W.
domain assumption The offspring distributions X_{i,j} are independent given the environment sequence.
Standard definition of branching process in varying environment.

pith-pipeline@v0.9.0 · 5419 in / 1299 out tokens · 27213 ms · 2026-05-10T19:08:44.787350+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

9 extracted references · 9 canonical work pages

[1]

Kersting, G., & Vatutin, V. (2017). Discrete time branching processes in random environment. John Wiley & Sons

work page 2017
[2]

Jagers, P. (1974). Galton-Watson processes in varying environments. Journal of Applied Probability, 11(1), 174-178

work page 1974
[3]

Agresti, A. (1975). On the extinction times of varying and random environment branching processes. Journal of Applied Probability, 12(1), 39-46

work page 1975
[4]

Kersting, G. (2020). A unifying approach to branching processes in a varying environment. Journal of Applied Probability, 57(1), 196-220

work page 2020
[5]

Korshunov, I. D. (2025). Branching processes in random environment with freez- ing. Discrete Mathematics and Applications, 35(4), 235-247

work page 2025
[6]

Pinelis, I. (2015). Best possible bounds of the von Bahr–Esseen type. Annals of Functional Analysis, 6(4), 1-29

work page 2015
[7]

Musielak, J. (2006). Orlicz spaces and modular spaces. Springer

work page 2006
[8]

Tanny, D. (1988). A necessary and sufficient condition for a branching process in a random environment to grow like the product of its means. Stochastic processes and their applications, 28(1), 123-139

work page 1988
[9]

I., Geiger, J., Kersting, G., & Vatutin, V

Afanasyev, V. I., Geiger, J., Kersting, G., & Vatutin, V. A. (2005). Criticality for branching processes in random environment. 12

work page 2005

[1] [1]

Kersting, G., & Vatutin, V. (2017). Discrete time branching processes in random environment. John Wiley & Sons

work page 2017

[2] [2]

Jagers, P. (1974). Galton-Watson processes in varying environments. Journal of Applied Probability, 11(1), 174-178

work page 1974

[3] [3]

Agresti, A. (1975). On the extinction times of varying and random environment branching processes. Journal of Applied Probability, 12(1), 39-46

work page 1975

[4] [4]

Kersting, G. (2020). A unifying approach to branching processes in a varying environment. Journal of Applied Probability, 57(1), 196-220

work page 2020

[5] [5]

Korshunov, I. D. (2025). Branching processes in random environment with freez- ing. Discrete Mathematics and Applications, 35(4), 235-247

work page 2025

[6] [6]

Pinelis, I. (2015). Best possible bounds of the von Bahr–Esseen type. Annals of Functional Analysis, 6(4), 1-29

work page 2015

[7] [7]

Musielak, J. (2006). Orlicz spaces and modular spaces. Springer

work page 2006

[8] [8]

Tanny, D. (1988). A necessary and sufficient condition for a branching process in a random environment to grow like the product of its means. Stochastic processes and their applications, 28(1), 123-139

work page 1988

[9] [9]

I., Geiger, J., Kersting, G., & Vatutin, V

Afanasyev, V. I., Geiger, J., Kersting, G., & Vatutin, V. A. (2005). Criticality for branching processes in random environment. 12

work page 2005