UT Wordmark Primary UT Wordmark Formal Shield Texas UT News Camera Chevron Close Search Copy Link Download File Hamburger Menu Time Stamp Open in browser Load More Pull quote Cloudy and windy Cloudy Partly Cloudy Rain and snow Rain Showers Snow Sunny Thunderstorms Wind and Rain Windy Facebook Instagram LinkedIn Twitter email alert map calendar bullhorn

UT News

Model Predicts Which Delinquent Credit Card Holders Will Pay

Research from the McCombs School of Business has identified a way to accurately predict which delinquent credit card accounts will repay an outstanding balance.

Two color orange horizontal divider

AUSTIN, Texas — Research from the McCombs School of Business at The University of Texas at Austin has identified a way to accurately predict which delinquent credit card accounts will repay an outstanding balance.

McCombs Assistant Professor Naveed Chehrazi and his co-author Thomas Weber from École Polytechnique Fédérale de Lausanne in Switzerland worked directly with a major credit card issuer to develop the Dynamic Collectability Score, which ranks delinquent account holders based on factors such as size of outstanding balance, mortgage status, past payment history, credit score, and external factors such as the performance of the stock market and the national unemployment rate.

The Dynamic Collectability Score continually adjusts as these variables change to provide a real-time prediction that is up to 50 percent more accurate than banks’ current scoring systems.

“Any new piece of information that comes in is going to change the prediction of the model,” Chehrazi says. “Each action that’s taken — from a collection phone call that goes unanswered to a partial payment that the bank receives — is factored in to revise, up or down, that person’s likelihood of future payment.” That information, in turn, improves the scoring system’s accuracy. No other current scoring system is capable of this, he explains.

Banks can calculate the likelihood that a credit card account will go into default, but once an account is delinquent, they only have a weak guess about who is most likely to pay back the debt.

Once an account is significantly past due, banks usually involve a third party collection agency to attempt to recover payment, but that strategy can be costly. Banks either sell off the debt for pennies on the dollar or pay hefty commission rates to recover only a small portion of the debt owed.

Banks need to know which accounts are worth spending money on — whether sending them to a collection agency, filing a lawsuit, or taking no action whatsoever — based on the likelihood of repayment and the amount they can expect to recover. Having that information would influence the strategy a credit card issuer follows for each account.

The Dynamic Collectability Score also helps banks better determine their credit risk capital requirements, or the amount of money they need to have in reserve to cover future unpaid accounts, known as Loss Given Default. Current estimation methods can result in projections that are either too low or too high, costing banks potentially a significant sum either through uncovered credit risk or increased cost of capital. Using Chehrazi and Weber’s Dynamic Collectability Score could ensure banks are adequately meeting their capital risk requirements established under the Basel II Accord — a set of recommendations for regulations in the banking industry.

“The credit card collection problem is very complicated,” says Chehrazi, “and current bank internal scoring systems are surprisingly poor in predicting repayment behavior, given the amounts that are at stake.”

Chehrazi and his co-author are in the process of expanding the Dynamic Collectability Score further so that banks can also calculate the optimal rate to pay collection agencies depending on each account’s repayment score.

The full study can be found here.