A system and method that performs a similarity calculation to identify insurance accounts having similar characteristics. The similarity calculation may be based on an existing insurance account or a synthetic insurance account that has user defined characteristics. The system and method may store account information for a plurality of insurance accounts that is normalized to a coordinate system. Parameters of a similarity account are identified and a similarity calculation is performed to identify a subset of the plurality of insurance accounts that match the similarity account. The similarity calculation includes calculating a distance between the similarity account and each of the plurality of insurance accounts and determining the subset of the plurality of insurance accounts that satisfy a predetermined similarity threshold based on the distance.