Definition:Bornhuetter-Ferguson method
📐 Bornhuetter-Ferguson method is an actuarial reserving technique used by insurance companies and reinsurers to estimate ultimate incurred losses by blending two inputs: actual reported (or paid) loss experience to date and an independent, a priori estimate of expected losses — typically derived from pricing assumptions, industry benchmarks, or the company's own loss ratio expectations.
🔧 The method works by splitting the ultimate loss estimate into two components. The portion of losses already reported is taken at face value from the loss development data. The remaining unreported portion is estimated by applying the a priori expected loss ratio to earned premium, then weighting it by the proportion of losses still expected to emerge based on historical development patterns. This dual-source approach makes Bornhuetter-Ferguson particularly valuable for immature accident years or long-tail lines such as general liability, professional liability, and medical malpractice, where early reported data can be sparse and highly volatile. Compared to the pure chain-ladder method, which relies entirely on extrapolating past development, Bornhuetter-Ferguson dampens the impact of unusually high or low early claims activity that might distort projections.
📊 Actuaries across the industry regard this method as a cornerstone of the reserving toolkit because it balances responsiveness to emerging experience with the stability of a prior expectation. Regulators and rating agencies expect insurers to use multiple reserving methods — and the Bornhuetter-Ferguson approach frequently appears alongside the chain-ladder and expected loss ratio methods in actuarial opinions and reserve studies. Its reliability, however, depends heavily on the quality of the a priori loss assumption; an inaccurate starting estimate will propagate through the unreported portion, which is why experienced actuaries stress-test inputs and triangulate results across techniques before recommending final carried reserves.
Related concepts