Setting a high downpayment or other requirements for mortgage loans would do far less to reduce defaults than it would to exclude borrowers with low incomes or from communities of color from homeownership, according to a new report from the Center for Responsible Lending and the Center for Community Capital at the University of North Carolina. The research findings focus specifically on the proposed rule for qualified residential mortgages (QRM) which implements parts of the Dodd-Frank financial reform law. Since the proposed rule came out, NHC and many others in the housing community have united around this very issue—that federal regulation should not set a 20% downpayment as a threshold for mortgage lending (see, for instance, NHC’s comment letter and past blog posts).
What’s new in this study?
- Directly addresses the tradeoff between reducing defaults and restricting access to credit. Proposed restrictions based on loan-to-value ratio (LTV), debt-to-income ratio (DTI), and credit score exclude a lot of borrowers to achieve very small reductions in the default rate. The study’s concept of benefit ratio provides a new way of evaluating the appropriate level, if any, of LTV requirement. For instance, the study shows that an LTV restriction of 97% (comparable to a 3% downpayment) provides more benefits in reducing default while excluding fewer borrowers than a 90% or 80% LTV.
- Relies on new data. The data set for the study includes data from loan servicers and from investor pools, which means the study could analyze more subprime and Alt-A mortgages (the more problematic loans, in other words). Data released with the proposed rule, in contrast, came from the GSEs and therefore skewed away from subprime loans.
- Starts from a baseline of product type restrictions. QRM is a layer of regulation on top of qualified mortgages (QM), which exclude many problem loan types such as negative amortization or exploding adjustable rate mortgages. The study therefore looks at the incremental change that QRM would make on top of the QM restrictions. Not surprisingly, QM alone goes very far to reduce defaults.