What it means
Insurance claims can take years to settle: a recent accident year may have only a few months of payments, while older years have a longer history. The chain ladder method uses the older years to estimate how much the newer claims may grow before resolution.
Data are placed in a development triangle, where each row represents a period in which claims originated, such as an accident year, and columns represent ages, such as 12, 24 and 36 months after that period began, so recent rows have fewer observed columns. An actuary can use paid losses or incurred losses, but the selection changes the interpretation.
Paid data track amounts disbursed, while incurred data include case reserves for known claims, and both may be affected by reporting and settlement practices. The method examines how cumulative claims changed between development ages in the observed rows, so a 12-to-24-month factor of 1.5 says that, in the chosen historical experience, a 12-month cumulative amount developed to about one and a half times as much by 24 months.
Factors are selected for each age interval and multiplied forward for an immature year, and an additional tail factor may be needed if claims continue developing beyond the last observed column. The projected final figure is called ultimate loss.
The Casualty Actuarial Society describes these age-to-age or link ratios and their use in moving from less mature losses to later development, and its research also studies when the simple chain-ladder projection can be biased, so a standard method should not be mistaken for a self-validating one. If a recent year has $2 million in cumulative paid losses at 12 months and the cumulative factor to ultimate is 2.4, the illustrative ultimate is $4.8 million.
The additional $2.8 million is not necessarily all unreported claims, since it can include known but unpaid claims and further development. Reserving labels require care too, because incurred but not reported, or IBNR, can be used differently depending on accounting and actuarial context, and the difference between ultimate and current paid loss is not automatically identical to IBNR alone.
The core assumption is that historical development patterns are informative for current claims, but a new policy type, changes in claims handling, inflation, legal developments or a catastrophe can break that relationship. Actuaries inspect the data and may select other factors or methods.
The method can be applied as a deterministic calculation, and statistical models can assess uncertainty, but the basic triangle and factors do not require a stochastic data-generation procedure. Reporting only one ultimate estimate hides the range of possible outcomes.
The point is to support an insurer's reserve judgment with a transparent view of claims history, so record selected factors, exclusions and adjustments, then compare past projections with later outcomes. A reserve is a current estimate that needs revision as claims develop.
In practice
Real-world examples.
Example
Historical paid claims rise from $1 million at 12 months to $1.5 million at 24 months. Their observed age-to-age factor is 1.5 before other years are considered.
Example
A recent accident year's 12-month paid claims are $2 million. Applying an illustrative 2.4 cumulative factor yields an ultimate estimate of $4.8 million, subject to the pattern's reliability.
Example
A claims department changes its settlement speed. The actuary reviews whether older paid-loss factors still fit the new process instead of applying them unchanged.
Formula
Calculation
Illustrative ultimate loss = current cumulative loss x selected cumulative development factor. If 12-month loss is $2 million and the factor is 2.4, projected ultimate is $4.8 million. The $2.8 million difference from paid loss includes several forms of future payment, not automatically only IBNR.
Building the factor from a triangle. Suppose two older accident years had cumulative paid losses of $1.0 million and $1.2 million at 12 months, rising to $1.5 million and $1.8 million at 24 months.
- Combined 12-to-24-month factor = ($1.5 million + $1.8 million) / ($1.0 million + $1.2 million) = $3.3 million / $2.2 million = 1.5.
- If the selected 24-month-to-ultimate factor is 1.6, the cumulative factor from 12 months is 1.5 x 1.6 = 2.4, matching the figure above.
- A recent year with $2 million paid at 12 months therefore projects to $2 million x 1.5 = $3 million at 24 months and $3 million x 1.6 = $4.8 million at ultimate.
The factors are invented and use only two older years; real triangles use more years, judgement about outliers and a tail factor.Case study
Seen in the real world.
Fictional example: An insurer's mature liability claims usually develop slowly after the first year. Noura builds paid and incurred triangles and calculates age-to-age factors for several origin years. A recent legal change increased settlement amounts, so she does not accept the raw older average without review. She tests a higher-severity scenario and compares an alternative reserving method. Management receives an estimate with assumptions and uncertainty rather than a single supposedly certain reserve.
Watch out
Common mistakes.
- Assuming past claim-development factors remain valid after inflation, legal or claims-handling changes.
- Calling ultimate loss minus paid claims only unreported claims, ignoring known but unpaid amounts.
- Presenting one deterministic estimate as the full range of possible reserve outcomes.
Questions
People also ask.
Why is it called a ladder?
Selected development factors carry immature claims through successive ages toward an estimated ultimate level.
Can it use paid and incurred data?
Yes. Both can be used, but the projections have different interpretations and sensitivities.
Does it guarantee the reserve is adequate?
No. It relies on selected patterns and assumptions that may fail when claims experience changes.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%Related