Closing the Underwriting Gap
As a result of this realization, Abraham, like the ABA's Johnson, says banks are beginning to recalibrate their credit models. But, he says, banks must go further, understanding that there's an underwriting gap. "There's a difference between the underwriting decisioning model and the borrower, business and market reality," Abraham contends. "The gap is between the ideal and the reality. Lenders need to start narrowing this gap with accurate loss prediction and less reliance on past prediction."
Abraham maintains that this gap exists in the data, model factors, sampling and model construction. He suggests that banks look to alternative forms of credit data, that models be designed to look farther into the future than they currently do, and that a hybrid model of sampling and model construction in which broader parameters are used to understand the customer more completely be adopted. "Lenders have to take control back from their models," he insists. "You need a hybrid model that uses sound principles and science."
Everyone agrees that banks and other lenders will need to be more careful going forward when evaluating potential borrowers, certainly in the subprime area, but also in the prime market. As an extension of this shift, says Robert Phillips, chief science officer and VP, research and development, with San Bruno, Calif.-based Nomis Solutions, since the credit crisis came to a head he has seen more investment in his company's price optimization solutions. To Phillips, however, banks' push to price products more strategically has a deeper meaning.
"What this says is that banks are looking beyond the world of price optimization and they're hungering to really understand their customers," Phillips explains. "They haven't done a good job of this in the past. So rather than just advertising better rates, there's a real desire to know the customers better. The winners will be those who can understand their clients at a deeper level. They'll want to take a lifecycle view of their customers, looking across products and time from the initial acquisition decision."