We were retained by a large U.S. insurer to assist in litigation to recover claims under its reinsurance policies. Based on a review of a sample of claims files by experts it had retained, the re-insurer alleged that our client overpaid and was therefore requesting amounts of coverage that were too high.
Our expert was retained to examine the validity of the claim file sample and evaluate the extent to which the sample could be extrapolated to the claims population.
Our team obtained summary characteristics of the entire claims population as well as the list of specific claims drawn by the opposing experts.
The opposing experts drew a stratified random sample using claim amount for the purposes of stratification. In general, the goal of stratification is to create strata that are more homogeneous (within strata) than the population overall.
Relatively homogeneous strata allow for smaller sample sizes to have sufficient predictive power over the population. While claim amount can be a powerful tool in stratification, when our team examined the sample, they found that the opposing expert had ignored systematic differences independent of claim amount that ultimately led to heterogeneous strata, the opposite of the desired outcome.
Because of this, the our expert was able to discredit the sampling results of the opposing expert on the basis of the sampling plan, selection methodology and sample size. Our expert submitted an expert report, sat for deposition, assisted counsel in developing cross-examination of opposing experts and testified at the arbitration.
Ultimately the panel of arbitrators found in favor of our client and ruled that the reinsurer was to pay out the policies.