Last week's announcement by the Federal Housing Finance Agency, Fannie Mae and Freddie Mac of changes in their representations and warranty framework is a step forward along a difficult path.
However, upon closer inspection, lenders remain on the hook for repurchase risk and must remain vigilant in assessing the quality of the processes and controls used to originate mortgages.
Since the financial crisis, billions of dollars have been at stake for the industry, the government-sponsored enterprises, and taxpayers to bring greater clarity to an otherwise gray area of the mortgage business.
To put the business problem into perspective, it is useful to think of a mortgage operation as a manufacturing enterprise that encompasses a number of related assembly line-like processes where a breakdown in any one of these functions can lead to product defects on the other side. In an effort to streamline the production process and improve throughput, management teams at times during the boom years took unnecessary shortcuts in the process leading to various malfunctions along the way.
One only has to look to various defunct companies as Countrywide Financial (where I worked as the bank subsidiary's chief risk officer from 2004 to 2007) and New Century to understand how such weaknesses in the loan manufacturing process could take place. The advent of automated underwriting and collateral valuation capabilities certainly improved our ability to triage credit and collateral risks, but they also precipitated an industrywide push to take on more risk through the loan manufacturing process.
Moreover, the introduction of various low-documentation programs, coupled with a variety of exotic mortgage products, further placed stresses on operational and governance processes ill-equipped to handle the volume and higher risk-layering that came about in the years leading up to the crisis. Ultimately, flaws in product design and the underwriting process have led to considerable claims of misrepresentation and fraud, leading in turn to repurchase demands.
The new GSE reps and warrants policy applies only to new deliveries. Underwriting deficiencies from the housing boom remain subject to repurchase review by the GSEs. The industry's movement toward plain-vanilla mortgage products since the crisis at least for now masks deficiencies in the underwriting process that remain to this day, despite industry and regulatory efforts to tighten a number of weak practices.
Responding to industry concerns over a lack of clarity with regard to the repurchase process, the GSEs outlined a number of specific delinquency thresholds for various programs that would trigger a repurchase. In another sensible move toward greater integration of the GSEs, Fannie and Freddie are aligning their lender contracts, thus reducing lender headaches over managing dual policies.
With the tying of repurchase requests to delinquency performance, lenders still have to worry about whether their underwriting processes are robust enough to guard against significant early payment default. It is widely accepted that material defects in underwriting that elevate the likelihood of a repurchase show up within the first 12 months after origination, in the form of early payment defaults. But defects may also manifest over a longer period. Thus, the new rules require a reasonably clean borrower payment history for as long as three years in some cases. This is where lenders can get tripped up by poor origination practices that lead to poorly performing loans and ultimately expose them to repurchase risk.
Lenders need to take a more methodical approach than ever before in measuring the propensity for repurchase and operational risks of the loan manufacturing process, adapting a combination of qualitative and quantitative tools usually reserved for evaluating mortgage credit risk. A repurchase risk assessment features a detailed, independent on-site, end-to-end evaluation of a lender's origination processes and controls including a review of sampled loans. A comprehensive survey, including hundreds of questions on all facets of the company's underwriting, operations, appraisal and broker management activities, quality controls and other critical business and risk management functions, is conducted in interviews with lender staff and from observations of field staff conducting the reviews. Responses to each question for a category are rated on a scale indicating the level of quality and effectiveness. These ratings are translated into risk weights based in part on their contribution to repurchase risk observed in the historical data. These risk weights are then aggregated across all categories evaluated in the process, resulting in a repurchase risk score, not unlike a borrower's FICO score. The score provides a consistent and objective rank-ordering of lender repurchase risk and in that fashion permits relative comparisons across lenders.
Large mortgage aggregators could greatly elevate their counterparty risk assessment by using such tools in determining eligibility requirements for their correspondent lender partners. Even the regulatory community, caught relatively flat-footed in sounding an early alarm about poor origination processes and controls, could benefit from the use of such tools by on-site examination teams. Armed with such capabilities, examiners can better target examinations to areas where pockets of emerging risks in the loan manufacturing process are observed.
The crisis has taught us many lessons about the need to better balance business and risk objectives. Shortsighted actions taken to rev up volume and/or increase efficiency turned out to have catastrophic consequences for the mortgage industry, as is painfully apparent today.
In an era of heightened awareness of the risks associated with poor origination practices, proactive assessment of risks across the continuum of the loan manufacturing process has never been as critical to long-term franchise viability as it is today. There is much to cheer about from the new GSE reps and warrants framework. But now is not the time to let our guard down in strengthening the underwriting process.
Clifford V. Rossi is the Executive-in-Residence and Tyser Teaching Fellow at the Robert H. Smith School of Business at the University of Maryland.