Assumptions and Mechanics of the Bornhuetter-Ferguson Method
Bornhuetter-Ferguson Method assumes that the reported or paid claims data does not contain enough information as to the amount yet to be paid. This deviates from the Chain-Ladder method where assumption of unreported or unpaid claims are solely based on the reported or paid claims data to date. Eric Brosius (1993) in his paper “Loss Development Using Credibility” described the Bornhuetter-Ferguson method as a credibility weighting between the development method (Chain-Ladder) and the expected claims method. Full credibility (Z=1) is given to reported or paid claims in the development method (Chain-Ladder) whereas no credibility (Z=0) is given to reported or paid claims in the expected
Bornhuetter-Ferguson Method assumes that the reported or paid claims data does not contain enough information as to the amount yet to be paid. This deviates from the Chain-Ladder method where assumption of unreported or unpaid claims are solely based on the reported or paid claims data to date. Eric Brosius (1993) in his paper “Loss Development Using Credibility” described the Bornhuetter-Ferguson method as a credibility weighting between the development method (Chain-Ladder) and the expected claims method. Full credibility (Z=1) is given to reported or paid claims in the development method (Chain-Ladder) whereas no credibility (Z=0) is given to reported or paid claims in the expected