---
path: /guides/chain-ladder-method
title: "The chain ladder method, explained"
description: "The chain ladder (development) method explained: development triangles, age-to-age factors, tail factors, and the seven-step process to project ultimate claims and IBNR."
section: Resources
priority: 0.6
changefreq: monthly
source_file: pages/marketing/seo/articleData.ts
---

# The chain ladder method, explained

The chain ladder (or development) method is the most widely used loss-reserving technique in property-casualty and health insurance. It projects ultimate losses and estimates IBNR from historical development patterns. Here is how it works, step by step.

## The core assumption

The chain ladder method assumes that historical loss development patterns are indicative of future development patterns. That assumption holds only when past loss experience accurately represents the future — so changes in product mix, claims handling, or the legal environment can make the results unreliable without adjustment.

## The seven steps

Following Friedland’s "Estimating Unpaid Claims Using Basic Techniques," the technique has seven steps:

- Compile claims data into a development triangle (rows = accident years, columns = development periods).
- Calculate age-to-age factors (link ratios / loss development factors) between successive valuation dates.
- Calculate averages of the age-to-age factors (simple, weighted, etc.).
- Select claim development factors using judgment after observing the averages.
- Select a tail factor to project from the latest age to ultimate.
- Calculate cumulative claim development factors by multiplying selected factors together.
- Project ultimate claims by applying the cumulative factors to the latest diagonal; IBNR is ultimate minus reported.

## Strengths and limitations

- Strength: simple, transparent, explainable, and based solely on historical data.
- Limitation: highly sensitive to data quality and to extreme values.
- Limitation: weak when external conditions change (e.g. inflation) unless explicitly adjusted.
- Often paired with Bornhuetter-Ferguson or the expected-loss-ratio method for immature years.