Ranking Analysis for Expectation of Binary Outcomes. A Bayesian Approach

The report delves into the estimation and grading of customer Probability of Default (PD) within financial institutions, highlighting the established use of a firm-wide master scale for internal portfolios. It addresses the ongoing debates regarding the comparison of grade PDs, average PDs across grades, and the observed Default Rate (DR). Additionally, it examines the Bayesian estimator for grade PDs, emphasizing its potential advantages over the traditional method of using simple default rates based on varying prior distribution assumptions.