As observers watch cash-strapped Fannie Mae and Freddie Mac -- the government-sponsored enterprises (GSEs) long deemed too big to fail -- being propped up by the government, many wonder how technology failed to save the agencies and the lenders supplying mortgages to them from being left holding the bag on so many bad loans.
Fannie and Freddie, which buy more than half of all U.S. mortgages, had combined debts of $1.5 trillion on the $5 trillion in mortgages that they own or guarantee as of mid-July. Their stocks had lost more than half of their value since the start of June, on news of $11 billion in mortgage losses -- echoing the sorry news from many banks lately.
Although technology alone could not have saved the industry from its current fate, analysts say, there are some technical/procedural lessons to be learned from the current crisis. These include the need for systems to better assess supporting loan documentation up front so lenders aren't forced to buy back loans sold to Wall Street or Fannie and Freddie when those loans transpire not to be as originally represented. The lessons learned extend to limiting the risks inherent in using mortgage brokers and better managing loan portfolios.
A decade ago, Fannie and Freddie developed automated underwriting systems and insisted that lenders use them, partly to ensure better credit quality. Indeed, early system output provided evidence of sounder loans. But then the rules of these rule-based systems were changed.
"In my mind it was not at all a failure of technology," says Walter O'Haire, senior analyst with Boston-based Celent, suggesting that technology only works when used properly. "There were systems in place that allowed lenders to set the parameters for terms on which they'd do the loan, and the GSEs had systems in place that allowed them to set terms as to what they considered conforming -- and those changed over time."
It was a calculated -- or miscalculated -- risk-return assessment, O'Haire says, "driven by the desire to make money."
Brokers of Bad Deals?
However, O'Haire does identify new software from firms such as Agoura Hills, Calif.-based Interthinx that flags scattershot applications by subprime applicants as valuable additions to lenders' technology resources. Subprime applicants typically make at least three times as many mortgage applications as prime candidates, often varying their stated income and liabilities, according to O'Haire. "These new [software] applications cut down on potential ... borrower 'misstatements,'" he says. Even simply using electronic origination, or e-mortgages, would help ensure loan applications have correct data, O'Haire adds.