The proposed Qualified Residential Mortgage rules have a better chance at harming the housing market than they have at addressing underlying weaknesses in the securitization process. They are an overreaction to the excessive risk-taking during the housing boom.
Regulators need to abandon the proposed approach at defining QRM-eligibility and implement a QRM scorecard that provides transparency to the market and allows for prudent but flexible risk-taking while also permitting well-qualified borrowers to participate in housing at reasonable costs.
In response to the lack of transparency and moral hazard issues surrounding mortgage securitization, the Dodd-Frank Act (Section 941) required regulatory agencies to establish risk retention rules for certain entities engaged in securitization activities.
These provisions require such firms to retain a portion of the credit risk associated with a securitized transaction unless the underlying loans qualify for exemption under a set of rules known as the Qualified Residential Mortgage provisions. Congress defined QRM-eligibility on the basis of individual risk attributes that are shown to reduce the risk of mortgage default such as strong credit history and downpayment. The proposed QRM rules, however, oversimplify the risk tradeoffs among borrower, product and collateral attributes in such a manner that is likely to raise borrowing costs and prevent a large segment of well-qualified borrowers from obtaining a mortgage.
To be sure, a number of the QRM restrictions reflect areas of product excess during the boom years such as negatively amortizing mortgages or piggyback 2nd lien mortgages that were used extensively in many markets as affordability products. It may be prudent to preclude them from QRM-eligibility, but it was extensive product morphing that contributed to the levels of credit losses observed during the crisis and not the presence necessarily of a specific risk factor. That is, taking otherwise standard mortgage products with stable historical performance and overlaying them with riskier attribute combinations created a set of new products with limited to no credit history from which to gauge future performance.
A more effective approach to introducing QRM-eligibility to the mortgage market would be to allow individual risk factors to trade off with each other in the form of a mortgage scorecard. For example, it is possible that a 90% LTV purchase money loan with a 720 FICO, 24% debt-to-income ratio, and fully documented income could have the same default risk as an 80% LTV, 690 FICO, 36% DTI loan that is also fully documented. Under proposed QRM provisions, a 90% LTV loan would not be QRM-eligible and thus be subject to higher costs if the loan were to be originated at all.
Standard mortgage underwriting practices allow for such tradeoffs and technological innovations in the form of statistically-based automated underwriting systems since the mid-1990s have demonstrated the power of such techniques to consistently and objectively evaluate multiple risk attributes together. Applying a QRM scorecard instead of the proposed QRM attribute-by-attribute restrictions provides greater precision in the measurement of credit risk than the proposed QRM rule that simply establishes limits on individual factors without regard to compensating factors that could offset the incremental risk of a factor.
The use of a QRM mortgage scorecard to assign loan-specific risk provides a more accurate and comprehensive indicator of mortgage credit risk. But a key issue is whether implementing such a tool is tractable across the entire mortgage industry.
The answer is that it is actually quite easy to develop and deploy such a model. The data, tools and resources to sharpen the pencil on mortgage risk for determining QRM-eligibility are available to federal bank regulators with ample risk management precedents existing in the mortgage industry and regulatory arena.
Implementing a QRM scorecard for determining QRM-eligibility would not only ensure an accurate depiction of mortgage credit risk, but would provide a large segment of well-qualified borrowers access to mortgage markets at reasonable costs.
Clifford Rossi is an executive-in-residence and Tyser Teaching Fellow at the University of Maryland's Robert H. Smith School of Business. He has held senior risk management and credit positions at Citigroup, Washington Mutual, Countrywide, Freddie Mac and Fannie Mae.