Searching for an Optimal Strategy for Identifying Files to Review for Fair Lending Analyses

Abstract

During mortgage transactions, there is a risk that some customers will be charged higher pricing based solely on group membership. Identifying such disadvantaged customers is a primary objective for banking regulatory agencies. Whatever strategy is used, there is an inherent tradeoff between resource allocation and reliability of conclusions. While the likelihood of identifying all disadvantaged customers should be high, the rate at which non-disadvantaged customers are identified should be low. This study uses Monte Carlo simulation to analyze how six different strategies compare on these two performance measures. The results show that the optimal strategy with respect to maximizing reliability and minimizing cost depends on the likelihood and severity of disadvantage. Further, none of the strategies are highly successful at identifying disadvantaged applicants or minimizing the number of non-disadvantaged applicants reviewed.