Equality and loan origination in mortgage lending
MetadataShow full item record
This dissertation focuses on the issue of whether inequality, specifically, gender inequality and racial inequality, exists in mortgage loan pricing. Prior studies on this inequality issue normally ignore the possibility that a borrower’s personal traits including gender and race might be associated with loan termination patterns, specifically default patterns and prepayment patterns. Failure to account for the possible association between a borrower’s personal traits and loan termination probability might lead to biased estimation of the impact of a borrower’s gender or race on a loan’s contract rate. Unlike prior studies, this dissertation examines whether a borrower’s gender or race impacts the loan contract rate beyond the extent to which it affects the probability of a borrower defaulting upon or prepaying a loan. Using data on the subsequent performance of each loan in the sample, a competing-risks loan hazard model is employed to investigate the degree to which a borrower’s gender or race is associated with loan default probability and prepayment probability. The resulting loan-level predicted default probability and prepayment probability are incorporated in a loan contract rate determination model to test whether a borrower’s gender or race has additional impacts on contract rate. The first essay focuses on gender inequality, and the second one concentrates on racial inequality. The results reveal that a borrower’s personal traits are indeed associated with loan termination probability. In terms of a borrower’s gender, the results show a female sole borrower tends to be more likely to default than borrowers in other gender groups; while with respect to race, African American borrowers and Hispanic borrowers are shown to be less likely to prepay than non-Hispanic White counterparts. After these associations are completely controlled for, the dissertation has found empirical evidence of both gender inequality and racial inequality in mortgage lending. Female sole borrowers are shown to pay higher contract rates than borrowers in other gender groups, while both African American borrowers and Hispanic borrowers tend to pay more for their mortgage loans than non-Hispanic Whites.