Home Statistical Dictionary Ruin Theory

Ruin Theory

The branch of actuarial mathematics studying the probability that an insurer's surplus (accumulated capital) is ever driven below zero by random claims, using stochastic process models of the insurer's income and payouts over time.

Stochastic ProcessesAdvancedFree

In Plain English

An insurance company collects steady premium income and pays out claims that arrive unpredictably, in random amounts, at random times. Ruin theory asks the central question every insurer's risk management depends on, given this randomness, what's the probability the company's surplus (capital reserve) ever gets wiped out entirely? Answering this precisely requires modeling both the arrival pattern of claims and their sizes as a stochastic process, and the mathematics developed for this specific actuarial question turns out to connect directly to random walk theory, Levy processes, and compound Poisson processes built elsewhere in this category.

Definition

The classical Cramer-Lundberg model represents an insurer's surplus at time $t$ as $U (t) = u + c t - X (t)$ , where $u$ is initial capital, $c$ is the constant premium income rate, and $X (t) = \sum_{i = 1}^{N (t)} Y_{i}$ is the aggregate claims process, a compound Poisson process with claim arrival rate $λ$ and i.i.d. claim sizes $Y_{i}$ . The central quantity of interest is the ruin probability $ψ (u) = P (U (t) < 0 for some t \geq 0)$ , the probability that surplus ever goes negative, starting from initial capital $u$ . Under the net profit condition ( $c > λ E [Y]$ , premiums exceed expected claims on average, without which ruin is certain), the Lundberg inequality gives the bound $ψ (u) \leq e^{- R u}$ for an adjustment coefficient $R > 0$ determined by the claim size distribution and the premium loading, and for many claim distributions an exact formula (e.g. the Pollaczek-Khinchine formula) is available. Ruin theory generalizes beyond the classical compound Poisson claims process to more general Levy-process-driven surplus models and to models incorporating investment income and reinsurance.

Formula

U (t) = u + c t - X (t), ψ (u) = P (U (t) < 0 for some t \geq 0)

Notation

u

—Initial capital (surplus at time 0)

c

—Premium income rate

ψ (u)

—Ruin probability, starting from initial capital $u$

R

—Lundberg adjustment coefficient, governing the exponential bound on ruin probability

Properties

The net profit condition is not a fine-print technicality, if average premium income doesn't exceed average claim outflow, the surplus process has a negative long-run drift and ruin becomes certain ( $ψ (u) = 1$ for every starting capital $u$ ), no matter how large the initial reserve, an insurer literally cannot outrun a losing proposition indefinitely.
The Lundberg inequality's exponential bound, $ψ (u) \leq e^{- R u}$ , gives insurers and regulators a concrete, quantitative way to translate a target maximum ruin probability into a required minimum capital reserve $u$ , directly informing real-world capital adequacy requirements.
Modern ruin theory extensions replace the classical compound Poisson claims process with more general Levy processes (allowing for correlated claim arrivals, heavier-tailed claim distributions, or continuous fluctuations from investment returns), broadening the classical Cramer-Lundberg framework considerably while keeping its core ruin-probability question intact.

At a Glance

Core modelCramer-Lundberg: surplus = initial capital + premiums - compound Poisson claims

Central quantitythe ruin probability

ψ (u)

Requiresthe net profit condition, premium income rate exceeding expected claim rate

Key boundthe Lundberg inequality,

ψ (u) \leq e^{- R u}

Last updated September 8, 2026← Back to the Dictionary