That’s certainly our first instinct in the world of credit lending: blame the borrower. But when it comes to loan defaults, the real failure may not be the borrower’s failure to repay, but the lender’s failure to manage risk.
To be clear, by risk management I mean all the activities and methods put in place to prevent a borrower from defaulting, thereby ensuring a profitable business for the lender, while avoiding driving the borrower into further debt.
And to get to the root of this failure, I’d like to look at three aspects of risk management that lenders can get wrong – and consider how we can get them right.
Credit scoring
The hallmark of risk management is credit scoring, a rating used to determine a potential customer’s creditworthiness and help guide the decision to lend. Or not to lend – particularly when a potential customer’s probability of default exceeds a certain threshold. When risk management is viewed through the lens of credit scoring, the decision to lend to a certain customer or not is the only lever for controlling risk.
Thankfully, recent developments in AI and access to increasingly rich data have seen the emergence of data analysis techniques that have improved our ability to predict credit risk. But as good as they are, these techniques are still far from giving us a crystal ball.
That means credit scoring alone is not sufficient for mitigating risk. We need to make use of other levers in product and service design – though each of which comes with their own challenges to get right.
Product design
Classic loan product design means thinking about the amount offered, collateral needed, repayment terms, and default provisions. This process, as applied by the rather conservative banking industry, has seen little innovation over the years, and continues to result in products that fail to manage risk. For example, the often excessive punitive financial penalties built into loan products are often counterproductive, both financially (by impossibly increasing debt) and psychologically (by deteriorating the trust relationship between the lender and borrower).
But now, the fintech ecosystem is shaking things up. By combining cutting edge technologies with fresh thinking in product design, fintech has helped new lending models emerge (or enabled previously-unfeasible models to become economically viable).
These new models include lending based on credit purpose (e.g. buying goods from a wholesaler on behalf of a merchant), repay as you earn (repayment from future earnings, e.g. for entrepreneurs on ecommerce platforms), and buy now pay later, which lately has been making inroads in Africa and generating a lot of headlines.
Another model that I think has a lot of promise is repay as soon as you can. It’s an optimised take on the classic loan model, built on better data, that provides incentives for early repayment and helps lenders distinguish excellent borrowers from merely good ones.
Whichever model you choose, I believe a well-designed loan product must be:
- Data rich – favouring generation of data with high predictive potential
- Short and recurrent – maximising renewal rates and the number of loans customers take each year to allow lenders to retain performing customers and adjust the eligibility amount of risky borrowers
- Tailored – with repayment terms adapted to the customer’s financial pulse (such as a sales cycle)
- Protective – preventing a spiral of over-indebtedness by rethinking how penalties are applied for late payment
Service design
While fintech has brought innovation to product design and increased the levers for managing risk, its models are weaker at addressing service. The reason for this is simple. By leveraging automation to scale, fintech products displace the traditional IMF or bank loan officer, putting at risk the kind of high-touch customer relationships so common in emerging markets.
As any loan officer knows, it’s only by forging close relationships that they can understand and monitor each customer’s specific situation. Good service design means helping borrowers understand what they’re committing to while mitigating any barriers to repayment. Here are some essential elements of good service design.
Repayment modalities
These are factors related to how, when and where the borrower can repay, as well as the kind of notifications they’ll get throughout the journey.
- Repayment logistics – To pay off the loan, does the borrower have to go to a branch or will an agent come visit them at their business?
- Digital services – Does the borrower have access to digital services, including the ability to automate payments?
- Timeliness – Will the funds be transferred on time? Will the borrower be notified about approaching deadlines or if the transaction isn’t validated?
Easy to understand terms
One of the most important enablers of good repayment is a clear understanding of the customer’s commitment. All too often, however, the terms and conditions are jargony, written in legalese, or are simply difficult to understand (sometimes intentionally so).
Planning for difficulties
Good service also means making arrangements in the event of payment difficulties, for example by rescheduling, postponing the due date, or refinancing. The product you choose must also be able to automatically identify these cases and match them with the most appropriate solution.
Human involvement
Every customer’s financial situation is uniquely complex and may entail challenging emotional issues that are best addressed by a human interlocutor. As much as possible, the customer should have access to a real person at some point in the customer journey.
Conclusion
When a customer defaults, it’s a double failure that results in debt for the customer and loss of profit for the lender. A healthy approach to risk management means doing everything possible to prevent this, through better scoring, product and service design.
There are a number of innovative approaches to each aspect of lending, but it’s rare to see solutions that focus on all three components at once. However, I don’t think it’s too ambitious to say that’s exactly what we should be aiming for: combining the best of fintech innovation with the benefits of the high touch customer service model of banks and MFIs. To make sure our approach to lending doesn’t simply accept failure as a hazard of the industry, but does everything possible to guarantee our customers’ success.