First-Time Defaulters Pose Technology Challenge for Banks

Counterintuitive as it might seem, there's a group of consumers banks should consider carefully for their profit potential: first-time defaulters.

Over the past two years, 22% of bank customers experienced a negative credit situation, and for 11% this was a new experience, according to the results of a survey from the Deloitte Center for Financial Services, part of Deloitte LLP. The survey was conducted in August, sampling 5,142 bank customers. The results were released Wednesday.

This group presents a number of technology challenges for the banks: How to identify them before they default and how to determine which ones might be profitable in the future.

Scoring technology and consumer data exist to do both, with vendors such as GDS Link LLC and Zoot Enterprises Inc. selling software that helps pull consumer data from numerous sources and helps make credit decisions in real time.

The question is whether banks will actually use it to create new customer models. "Banks are gun-shy once they have a default population. If they dip their toes back in these waters, it is fraught with dangers," said Andy Laing, a senior manager in Deloitte's banking group. Some of "these customers can provide value and a great relationship going forward, and there is a lot of pent-up demand on the consumer side, and [banks] could profit from this," he said.

Typically, banks use information from a limited universe of data sources, such as credit bureaus and FICO scores, to predict customer behavior. But the economic crisis and regulatory focus on risk has put more emphasis on expanding the sources from which banks pull, experts said.

Banks should look to data that casts a wider net to determine other causes for defaults and delinquencies, Laing said. For example, demographic data might reveal that a customer lives in a particularly hard-hit region, such as Florida or Arizona. Similarly, banks could pay attention to the kind of product that caused a default, such as an adjustable-rate mortgage, which a first-time borrower may not have entirely understood.

The security vendor GDS Link helps banks create quick, customized profiles of consumers by tapping more than 30 traditional and nontraditional databases. "You need a good portfolio management system where banks can take all their portfolio data across each line of business and each system in each line of business," said Paul Greenwood, the Dallas company's president. Such systems should continually blend in data from alternative credit sources, including databases that track payday loans, title loans for automobiles, employment history and salary.

Taking a slightly different approach, Zoot Enterprises helps banks make real-time decisions, including at the point of sale, by tapping more than 40 alternative databases such as those that monitor 8(a) housing loans. This data is blended with standard credit bureau information.

"We do this with retailers every day when consumers are in the check-out line, or a bank representative is screening them for a credit card offer," said Dennis Dixon, the president of the Bozeman, Mont., company.

But there are still challenges in determining a customer's future profitability. Banks' customer data has traditionally been siloed by product line. While it might be easy to get a clear picture of how a customer fares with credit cards, home and automobile loans may be an entirely different matter.

"Additional data, beyond credit bureau data, can add insight on the customer's financial position, including changes in housing type, automobile asset quality and secondary financial needs, such as payday lending usage," said Brian Riley, a senior research director at TowerGroup in Needham, Mass. "These factors can supplement thin bureau files or even more importantly, broaden the data points used in credit decisioning."

Another problem is lack of consistent data. "Between lines of business, the same piece of data might mean different things," said Tom Johnson, the vice president of product development for Zoot.

A bigger issue is that banks' models for measuring risk can take up to two years to create, making them outdated in many cases, experts said. "One of the services we provide banks is the ability to do simulations and to shorten the loop for getting new models into production," Johnson said. GDS Link can also help banks lower their model-building time to a matter of months, it said.

The Deloitte survey also pointed out that 41% of first-time defaulters had a bad impression of their bank, and 63% said they were not at all likely to ever borrow from the bank again. That, observers said, makes it imperative that banks handle default situations correctly.

One way technology can lend a hand is through collectors' portals like Web Promises, offered by CGI Group, said Christine Pratt, a senior analyst covering lending for Aite Group LLC in Boston. Such portals unify a bank's responses to negative credit situations, minimizing the possibility that customers will run into harsh collectors or call center representatives, she said.

Regardless, banks need to use loan portfolio management analytics, predictive analytics, and behavioral modeling and optimization much more intensively than they do today, said Kristin Moyer, the research director for the banking and investment services industry advisory at Gartner in Stamford, Conn. "Having a high percentage of first-time defaulters in your customer base poses risk, but also opportunity," she said. "Banks need to use loan-portfolio management analytics to segment this group."

For reprint and licensing requests for this article, click here.
Bank technology
MORE FROM AMERICAN BANKER