The U.S. and global economies ended the year 2019 on a positive note. After a decade of expansion in the job market, the U.S. unemployment rate was under 4% and the stock market had reached an all- time high. Three months later, the world was in lockdown. In the U.S., after several years of steady growth, the GDP fell by 4.8% and unemployment was on its way to historic highs. Economic indicators suggest the possibility of a recession. How can financial institutions better manage credit risk in this uncertain economic climate?
To discuss the use of artificial intelligence (AI) in the assessment of credit risk and prevention of delinquency, Brian Riley, Director, Credit Advisory Service at Mercator Advisory Group and Amyn Dhala, Vice President for AI Express within Mastercard’s Cyber and Intelligence Solutions Division joined host Samantha Maloney in the Mastercard InConversation Series webinar, Assessing Today’s Credit Risk and Mitigating Tomorrow’s Delinquency with AI.
Household Debt and Credit Risk
Household debt, led by mortgages and credit cards, exceeds $14 trillion. Lenders are increasingly concerned about the rising debt level and the economic impact of the global pandemic. It “is increasingly important to monitor and identify borrowers who are finding it difficult to pay down their debt and work with them to manage the risks and the consumer needs,” stated Riley.
Traditional approaches to loss mitigation are typically reactive. Accounts are flagged only after payments are missed, which may be too late for lenders to address the underlying issues, collect on the already delinquent account, and retain the customer.
Proactive solutions identify problematic accounts before they become delinquent. By allowing for the earliest possible intervention, banks are able to collaborate with customers to find solutions that benefit both lenders and borrowers alike.
“Banks are looking to leverage technology and specifically artificial intelligence to achieve the twin full objective of … reducing/optimizing the credit risk and continuing to provide [a] good customer experience,” said Dhala.
Benefits of AI in Credit Risk Management
- Improving the customer experience through personalization allows banks to “provide the optimal experience to that particular customer at that point of time,” explained Dhala.
- The ability to predict delinquencies before they occur prompts early action to reduce credit losses and associated collection charges. A good AI model “can detect delinquencies as early as 12 months ahead, assuming that we have the right data sources in place and [are] able to build a robust model,” noted Dhala.
- Managing risk across the customer lifecycle by constantly monitoring and evaluating customer behavior allows banks to not only mitigate potential losses, but also extend credit to meet customer needs.
- Leveraging data across an organization improves prediction accuracy in real-time. This ability facilitates the prediction of accounts that are at increased risk of delinquency, the identification of potential fraud, and the reduction of false transaction declines.
AI Express
The benefits of AI in credit risk management are even more relevant in the current economic climate. Many companies are looking to incorporate AI into their risk management strategies, but lack the experience and capability to create accurate and reliable models. Others who are already using AI may be looking to improve upon their existing techniques. With AI Express, Mastercard is helping companies develop AI models that meet their specific needs.
AI Express is a two-step process. Step one combines an organization’s business experience and prior analytics with Mastercard’s AI technology and expertise to design a superior model. In step two, the newly designed model is deployed.
Over the course of six to eight weeks, Mastercard guides the model development process through six stages, resulting in an optimized model that can provide personalized, accurate results specific to each individual.
- Business understanding: Work with clients to isolate the specific use case they want to address.
- Data understanding: Gather and validate data from all available sources.
- Data preparation: Select data best suited to specific client needs.
- Modeling: Choose appropriate modeling techniques and build a customized model.
- Evaluation: Run through the model, evaluate results, and determine the next steps.
- Deployment planning: Review deployment options and create a plan.
Conclusion
Artificial intelligence offers the most effective credit risk management tools for financial institutions. Successful AI models gather, evaluate, and learn from vast amounts of data giving them the ability to personalize risk assessment, adapt to new information, and scale as well.
“We are amidst change, and it is unprecedented change, and it’s a good time to look at many of the ways you approach particularly … the credit risk management side. And I think that from what we’ve seen today, this [AI Express] offers a good option,” concluded Riley.
