Outstanding claims reserving have become most of the time Best Estimate whereas they used to be appropriate. These reserves should now be equal to the best estimate of the cost of the claims not yet settled and not yet reported. Even if new reserving methods have flourished recently, the most typical ones, Chain Ladder and Bornhuetter Ferguson, remain by far the most popular methods in the actuarial world.
Choice of reserving methods, data construction or determination of the underlying hypothesis are often in the hands of the actuary and their judgment, usually called expert judgment. The leading role that judgment plays can explain why two actuaries with the same data could obtain two different Best Estimates. The assurance of a high quality expert judgment, is therefore necessary in order to ensure the quality of the estimate itself.
As anyone else, the actuary may suffer from the effect of cognitive biases which could damage the quality of their judgment. Anchoring, status quo or representativeness biases are some of the many threats to providing the best judgment and with it the best estimate. Identifying those biases and experimenting in the similar conditions to reserving is therefore necessary.
This article aims to introduce the results of a statistical study made of mock reserving cases and delivered to a group of actuaries in which they unconsciously faced cognitive biases. The results will allow to state if these biases truly have an influence and to give a first measure on the Best Estimate measure.