Consumers have been taking advantage of low interest rates to borrow money, but recent trends show that an increasing number of borrowers are having trouble keeping up with their payments. Consumer debt and delinquency rates have been on the rise for the past several years, with U.S. consumer debt topping $14 trillion in 2019. The current economic crisis will inevitably accelerate this trend, further increasing the risk for lenders.
Delinquency is tracked in 30 day increments (30+, 60+, 90+ …). The more delinquent an account becomes, the more difficult it is to collect. Typically, after 180 days, delinquent accounts are written off as bad debt and sold to third parties for collection; the outstanding amounts are no longer reported as assets on balance sheets.
Debt Collection is Costly for Lenders.
When payments are missed, creditors contact customers to try to collect past due balances. Initial contact is generally made by the lender, but in time the account may be turned over to a third‑party debt settlement company (DSC). Either way, debt collection is a costly endeavor. Furthermore, attempts to collect on delinquent accounts can alienate customers and lead to attrition, a situation that creditors would like to avoid because the cost of acquiring new customers is even higher than the cost of retaining existing ones.
In an effort to mitigate their losses, financial institutions and credit card issuers have developed numerous programs aimed at reducing collection costs and retaining customers including: re-aging delinquent accounts, forbearance, debt settlement, and credit counseling. These traditional loss mitigation programs share one common denominator: they are all reactive. Since no action is taken until payments on the account cease, it is significantly less likely that collection efforts will be successful.
Proactive Solutions
Lenders need proactive, personalized solutions. They need to assess risk in a way that allows them to predict delinquency before it happens and to initiate action while there is still time to prevent avoidable losses.
To meet their needs, lenders are looking to artificial intelligence (AI) and machine learning (ML). Machine learning is an application of AI that allows computers to analyze and learn from vast data sets to make intelligent inferences, and improve its performance over time. AI and machine learning technology can be used to compile and assess data from multiple sources in real time, including credit card use and online banking transactions, to create a behavioral profile and snapshot of a customer’s current financial situation.
By flagging at-risk accounts and alerting lenders to the likelihood that a customer is headed toward delinquency, AI provides creditors with a valuable opportunity for early intervention to reduce default losses.
Not All AIs are Created Equal
AI models are only as good as the data on which their assumptions are based. Higher quantity and quality data lead to more reliable predictions. Brighterion, a Mastercard company leverages its smart agent AI technology, proprietory modeling techniques to create highly personalized and highly accurate predictive AI models.
By using a more accurate AI model to monitor accounts, changes in customer behavior are evaluated in real time to determine which accounts are at risk before the first missed payment. Knowing that a customer is likely to default on payments in the near future allows lenders to intervene at the earliest opportunity. With customer specific insights gleaned from AI, delinquencies can be reduced and collection strategies can be personalized for individual situations, providing a more positive consumer experience to protect the customer relationship and reduce lost revenue.
Brighterion has developed a fast, flexible credit risk analysis system that can readily adapt to evolving markets and customer profiles allowing lenders to predict and prevent delinquency despite rising debt and an uncertain economy.
To learn more about how Brighterion AI can help prevent credit delinquency, You can download the complimentary whitepaper below.
