Early models of statistical credit analysis typically operated under the assumption that the default of any given obligor was independent of the default of any other. This simplification made the models more tractable but less realistic, as it did not account for potential correlations between defaults (e.g., economic downturns affecting multiple obligors simultaneously).