This story appears in the January 2009 issue of Cards&Payments.
Financial institutions will face increased scrutiny for years to come. That means credit card issuers will need to explain underwriting decisions in more detail, both to investors and to regulators, says Andy Jennings, chief research officer at Minneapolis-based Fair Isaac Corp.
Clear explanations of complex decisions can be difficult. Card issuers base credit-underwriting and clearing decisions on analysis of data that is much more extensive than data issuers used in the past. That can lead to "a decision tree and thousands of end notes human beings can't make sense of," Jennings says.
Decision trees are reports, often lengthy, showing how a multitude of data points led to a particular credit underwriting decision. Fair Isaac launched a product called Strategy Designer earlier this year that translates those decision trees into a more-understandable format, partly by stripping away any repetition of data.
Speed of development and deployment is a common goal among developers of credit card risk-management products and of the financial institutions that use them. But speed has become more important during the recession and will continue to be a priority post-recession, according to Jennings.
"We have all these bright people, either Fair Isaac scientists or scientists the big issuers hire themselves, and they take six months to get the model [completed and deployed]," he says. "In stable times, that may not be as much of a problem. But since there's more uncertainty predicting credit risk today, that time lag becomes more important. That's a driver now and will prevail when we come out of this thing."