The current trends in financial markets are rapidly changing, and lenders are facing many challenges as delinquency rates climb in both consumer and SME lending. When borrowers lose jobs, have medical emergencies, or have inconsistent cash flows, they frequently end up defaulting on payments. All of these events negatively affect not just the lender’s portfolio but the borrower’s ability to pay. Traditional collection methods begin to operate after the payment has been missed and stress the borrower with continued aggressive follow-up methods after the fact. New AI-driven early prediction of delinquency will allow lenders to begin supporting borrowers prior to their failure to pay on time. By identifying risk before the borrower becomes delinquent, lenders can provide proactive, customized support to help borrowers stabilize their financial lives, thus improving lender recoverability and operational efficiency.
The introduction of predictive analytics tools enables lenders to manage risk differently, predicting the likelihood of an account defaulting prior to the date of default, allowing lenders to alleviate NPAs, better recover from potential losses, and provide more efficient and focused support to those at-risk borrowers. Rather than using expansive and inefficient marketing strategies, lenders now have the capability of targeting an intervention to those borrowers who most need it, resulting in decreased operational expenses and improved overall portfolio performance.
Borrowers also benefit greatly from these payment options and/or reminders, as they help the borrower avoid late loans by offering timely nudge reminders. These reminders reflect a borrower’s own unique circumstance, such as income variability due to gig work, or shifts in cash flow seasonally. Reducing the number of late fees, along with the surrounding stress of debt collection calls, also helps to lessen emotional stress related to debt collection. More importantly, by creating a relationship of empathy between the lender and borrower, they can be more open and communicative with one another instead of being afraid to communicate or work together.
The static nature of credit scores based on traditional or their use in legacy risk models creates a disconnect to the true financial life of the borrower. Sudden disruption in income, unemployment, unanticipated costs, etc., will not be identified as issues until after they occur i.e., after the borrower has defaulted on their loan payment(s). Also, when loan portfolios are manually reviewed, and borrowers are contacted reactively (e.g., when they have defaulted on their loan payments), there is no opportunity to constructively resolve situations prior to reaching this peak level of stress.
The data silos also separate the knowledge of how borrowers behave from the lender. Lenders cannot send their message to borrowers using unified real-time information about transactions, communications and income fluctuations. As a result, lenders often send disconnected, impersonal messages to borrowers, which many borrowers reject. This subsequently reduces the positive interaction potential between lenders and borrowers and has a negative effect on recovery rates.
Artificial Intelligence (AI) uses a unified dataset that captures operational, financial and behavioral signals to develop predictive analytics that can identify early warning signals of delinquency. AI utilizes transaction data, repayment history, job changes, banking statement data, device and application usage, and customer service interactions (e.g., via phone, email, text message & field) as well as behavioral events such as changes in repayment behaviour, bounce rate, declining income, spikes in credit card utilisation, or changes in communications engagement to feed complex machine learning predictive models.
AI uses this information to continuously update borrowers’ risk scores as the signals they receive change over time. This allows for the early identification of silent struggling borrowers who have not yet missed any payments but show signs of financial stress. The AI also can identify which of these borrowers only need a gentle nudge, and which require more substantial payments or payment restructuring solutions. As a result, lenders can anticipate potential delinquencies on loans before they become due and take action to solve these issues with compassion rather than react with aggression at the last minute.
With the use of early risk scores, lenders are able to create tailored engagement strategies to suit different types of profiles for borrowers (i.e. salaried employees, gig economy workers, small seasonal business owners). Empathetic message communication that conveys empathy such as “We know this month may be tough; here are a few ways to make payments easier” helps the lender receive positive responses and cooperation from affected borrowers. Providing options to adjust EMI payments, depending on individual borrower’s cash flow cycles and/or to facilitate temporary restructures to reduce further delinquencies will benefit the borrower.
Customer personalisation helps enhance a positive experience for customers, but also greatly increases the likelihood of successful interventions and decreases the cost for collections. When applying AI technology for optimising timing, messages and communication methods, this approach will auto-arrange that all outreach will be sent at the best time and in the manner in which borrowers respond best.
Through AI-driven early delinquency prediction, lenders can provide a supportive and proactive collections process for borrowers while achieving lender goals. Therefore, lenders that embrace this technology will reduce their NPAs, decrease collection costs, and establish healthier portfolios for their customers. Additionally, borrowers will receive tailored support to eliminate late fees and the public perception of defaulting on a loan.
As borrowers’ financial situations continue to become more complicated and diverse due to the increasing complexity of credit markets, lending institutions are moving towards empathetic AI-driven early warning systems. These systems will enable lending institutions to establish trust-based and resilient relationships with their customers while pursuing sustainable business growth.
Views expressed by: Neeraj Kulkarni, COO and Co-Founder, Collectedge
