Table of contents
Getting a loan on the book is only one part of the equation, monitoring it until the installments become due and seeing it through until it is fully paid and closed is a whole new ball game. However, for most lenders, once a loan is approved, it goes out of the picture and becomes the problem of the collections team. This fractured view of loan as a series of steps to be taken in silos by various teams is probably why lenders see high NPAs and contribute to the current delinquency mess in India. At the same time, lenders have long employed rudimentary collections strategies that use static parameters such as due date and invoice value, making collections processes reactive rather than proactive.
More importantly, most lenders view underwriting and collections risk through the same lens. But effective credit management goes beyond underwriting good loans to include performance measuring, modeling, and forecasting.
Treating risk distinctly at every stage of the loan life cycle, especially during collections, is a key component of sound loan management practices.
Why treat collections and underwriting distinctly?
Lenders have long assessed collections risk by using models resembling those used for underwriting consumer credit. Similar treatment of the two falls short on several fronts.
Firstly, during the entire span of the loan tenure, a borrower is likely to touch several peaks and troughs of risk trends. For instance, a company with strong financials at the time of loan approval can find its reserves depleted at the time of repayment due to factors ranging from environmental disasters to a global financial crisis. That is why closely monitoring the entire life cycle of a loan account, from underwriting to collections, and deploying early warning signals is of paramount importance.
Secondly, with rising delinquencies and resource limitations, lenders need better segmentation and a narrowed down list of at-risk borrowers for prioritized collections operations. Analytics help improve segmentation efforts and enable tailored contact strategies.
Thirdly, collections-specific risk models also factor in data during different stages of delinquency (30, 60, and 90 days of delinquency and post-default curing period), which help design customized interventions, thereby, improving collections efficiency.
All this is best achievable through contextualized risk assessments enabled by technology. A strong AI- and ML-based loan management engine that tracks portfolio health through dynamic indicators can help identify loans that are likely to be stressed and remedy delinquent accounts before they lapse into defaults. Tech-aided, contextualized risk assessment models are increasingly helping lenders navigate amid chaos and meet their strategic objectives.
Why is contextualization important?
Debt collectors tend to overlook the fact that delinquency often stems from simpler, mundane reasons: borrowers forget their due dates, face a temporary liquidity problem, or simply find repayment formalities overwhelming. In many cases, a delay in repayment does not mean unwillingness to pay. And pushing buttons by flexing muscle or resorting to repetitive phone calls result in needless unpleasantness.
Instead, lenders can foster positive customer experience even at the collections stage by using technology and advanced analytics. A collections-specific portfolio monitoring model driven by machine learning, analytics, and prioritized segmentation would make for a multi-dimensional matrix that factors in a variety of dynamic parameters. The insights from this model enable lenders to determine their contact strategies — from deciding which communication channels to use to guaging the tonality and level of urgency to be deployed in the communication.
This nuanced approach has proven to create significant value. A McKinsey study states that it can reduce NPAs by 20-25%, increase resolution rate by 25%, cut down collection cost by 15%, and increase customer engagement 5X.
How FinBox CollectX prevents delinquencies from becoming bad debts?
At FinBox, we design collections-specific models that leverage behavioral analytics to predict the probability of customers’ prolonging their delinquency period or defaulting and help lenders prioritize collections operations. With our 16 million new-to-credit borrower data and over a million underwritings done every month, we have the right models and analytics that can help lenders predict portfolio quality with more accuracy.
In addition, user-specific analytics are arrived at using real-time data on financial activity of borrowers. Our state-of-the-art early warning system raises red flags at different stages of the loan life cycle — prior to installment due date, during different phases of delinquency and curing period.
The illustration below gives you a glimpse of how such AI-led collections models enable a layered approach to tackling delinquency in comparison to traditional methods.
The next-generation collections environment will be powered by real-time data and advanced analytics. The transformation has already begun. Technology is enabling lenders to rise to the challenge of delinquency by equipping them to renew collections strategies in real-time for different customer segments. With insights into borrowers’ financial behavior on a day-to-day basis, lenders can proactively manage risk and check potentially crippling credit losses in a future downturn. A proactive and nuanced approach becomes the most important, especially at the collections stage of a loan life cycle. After all, collections, the last leg of the race, is what makes or breaks a lending business.