By Arnab Gupta and Phillip Riese
Panicked by the prospect that credit card defaults could rival those of the mortgage meltdown, banks are initiating aggressive recovery programs against borrowers at an unprecedented rate. Our analysis says there is a better way.
All delinquent borrowers are not created equal, differing in both ability and willingness to pay. Lenders must determine where a debtor stands to best define the collections approaches for individual customers, a calculated strategy to guard important client relationships while maximizing debt repayment. Unfortunately, most banks work with credit models designed for more stable times, relying on data that goes back several years. Today, a customer’s situation can change dramatically in only a few months. Over the past year, we have been working with major banks and financial institutions to develop credit models that are finally nimble enough to incorporate rapid changes in a customer’s situation to effectively assess credit-worthiness in today’s dynamic economy.
Opera Solutions has increased predictive accuracy 15 to 25 percent by changing the analytical methods we apply and data we use. For example, neural networks and nonlinear techniques that look at relationships among a larger number of complex variables allows better segmentation of the customer base than previous methods. In addition to economic factors like changes in employment status or real estate values, we overlay variables such as transaction data by size, type, and category; utilization rates; and relationship factors like tenure, number of products, and profitability to create a picture of the customer’s true value. These models distinguish between consumers who can manage increased credit lines from those who should be watched more carefully. Moreover, our advanced, holistic modeling allows us to determine the type and level of intervention that would be most appropriate for each client. Although these new models make distinctions very finely at the individual consumer level, we have identified five basic types of borrowers. Being able to tell which is which can make a difference of as much as $200 to $500 million to a large lender.
So where do borrowers fit?
False Negatives: Customers who can handle higher credit lines, and deserve them, but who have been misidentified by obtuse models that can’t distinguish among different types of borrowers. Our analysis shows that two-thirds of recent credit actions have impacted what are generally the most financially responsible consumers. Instead of alienating these customers and exposing the business to franchise risk, the appropriate strategy would be to nurture them, potentially with greater credit lines, and ultimately win market share, long-term loyalty and profits.
On the Edge: Good customers whose behavior sends up a sudden red flag. An adverse action against a good customer, particularly one not in delinquency, is a breach of trust that’s difficult to repair. It drives customers to competitors, ruins brands, and kills performance, so it’s critical to be able to identify good customers who are going to pull back from the brink without draconian measures.
Recoverables: Responsible borrowers who would repay if they could. Spurred by the economic recession, this group is larger than ever before, now representing 20 to 25 percent of defaults. A smart approach applies customized programs that nurture good-customers-gone-bad back to financial health. We find that handling these customers with sensitivity delivers more than two and a half times the money as heavy-handed approaches.
Dry Holes: Among customers who wind up defaulting, 34 percent will never pay anything. The key is to ensure these customers don’t take up scarce internal collection resources. These cases need to be closed quickly and their files moved to external agencies. Doing so can free up 10 percent or more of a bank’s collection capacity.
Just Need a Nudge: Twenty percent of defaults will cure with minimal intervention. Don’t use expensive channels like a personal collection call when a letter or text message will be just as effective.
The key is to identify where each individual fits – but in a sea of defaulting credit card customers and a fast-changing, complex economic environment, it’s not easy. What can help: more sensitive and flexible segmentation models incorporating very recent behavioral data, visibility into customer payment patterns over the long term, and good old-fashioned business judgment. Knowing what type of debtor lenders are dealing with can point the way to detailed contact and collections strategies that improve recoveries, foster loyalty, and support higher profits in the long term.
Arnab Gupta is the CEO and Phillip Riese is the Senior Advisor of Opera Solutions, a global management-consulting firm helping the world’s premier companies quickly achieve rapid, tangible, sustained profit growth through its proprietary “Total Data, Total Insights, Rapid Action” approach. For more information, please contact Laura Teller, Chief Communication Officer, firstname.lastname@example.org; 646-520-4338.