The value at risk metric employed by every major banking institution offered the appearance of safety until taxpayer bailouts were the only option to save the financial system. Today, regulators are examining historical factors to determine what defines a “safe” mortgage in order to calculate capital adequacy and risk retention requirements.
Last year, lenders declined over $1 trillion in mortgage applications that would have helped the housing industry that is entering its fifth year of depression. Recently, FHA delinquencies are spiking alarmingly even as borrower credit scores have hit historic highs.
Long after the 2008 meltdown, banks and agencies remain heavily invested in credit models that rely on data that doesn't inform and analysis that doesn't predict. New thinking about the changing dimensions of risk is needed to restart the flow of consumer credit.
Lenders stop making reasonable loans when they lose confidence in their ability to predict losses.
The mortgage industry especially is facing the uncomfortable reality that traditional models and metrics, like credit bureau scores, are far less relevant in an economy dealing with job loss, income instability and negative home equity. For example, an emergency room nurse in Dallas with a credit score of 660 is a better mortgage credit risk than a construction foreman in Naples, Fla., with a credit score of 740. The nurse will be better able to sustain income in a high demand profession, working in a stable local economy where property values are still largely unaffected by the housing collapse.
Past credit performance still matters of course, but not as much. Steering by gazing through the rearview mirror doesn't work when the road begins to make unexpected turns.
The obvious answer to the housing credit crunch is for government and banks to make more mortgage loans. This requires moving beyond credit bureau histories, loan-to-value calculations and debt-to-income ratios. Our firm recently tested a new default risk evaluation method on a $40 billion portfolio of 100,000 mortgage loans for one lender that demonstrated the importance of borrower job and income stability. The findings showed that the underwriting process passed on $2 billion in good mortgage loans that could have been made.
Further, this approach was 34% more predictive of mortgage defaults than a credit bureau score. Unfortunately, traditional consumer credit scoring models don't evaluate the new drivers of credit risk. A borrower's job, income and home location matter most in an economy featuring 16% structural underemployment, widespread income insufficiency and ongoing home value erosion.
Calcified thinking about mortgage credit can be tied directly to the position of the GSEs in the housing supply chain and their role in setting “conformance” standards. These factors were institutionalized to support securitization in the early 1980s and then automated in underwriting platforms deployed in 1995.
Today the GSEs are stuck in the outdated processes and infrastructure that helped accelerate and spread global contagion, encouraged to tighten their “lending standards” and strangling the housing market.
Reliance on old systems based on old thinking is everywhere in the mortgage life cycle. This has been especially evident in the loss mitigation catastrophe. Delinquencies remain high, 50% of loan modifications are failing, and 4 million to 6 million more foreclosures will be dumped on already distressed markets.
Underwater homeowners can't move to new jobs, slowing economic recovery. Housing is trapped in a vortex of banks not making or modifying loans and consumers being denied credit to buy or unable to sell homes. Our consumer economy will struggle until we begin to think differently about new solutions to break the negative credit cycle.
Tools designed for the pre-Internet era are not capable of dealing with the nature and scale of this century's problems. The inability to confront housing finance reform has locked the industry and their regulators into a “change we can't conceive” mindset from loan origination to asset sales to REO disposition. The reluctance to test new ideas, adopt new credit models, and upgrade to latest technologies has resulted in creditworthy consumers not receiving loans, housing markets not recovering, and jobs not being created.
Fostering innovation based on new thinking about mortgage credit risk is essential to extending more credit, unleashing economic growth and restoring the American dream.
Suresh Annappindi is the founder and CEO of Scorelogix LLC, a risk modeling and predictive analytics company founded in 2003 and headquartered in Delaware.
