Co-running a fintech startup gets you a lot of (well meaning) questions. We’re still in our early days, but the most pressing questions we get revolve around the number of loans we’ve disbursed or size of the book the size of the portfolio we’ve originated. It’s understandable that the questions are along these lines since most media reporting, alongside fundraises, focus on these metrics as an indicator of success.
But to understand of the health of a lending business, this is a grossly inadequate, and insufficient, metric of evaluation.
Money: The Product That Sells Itself
Lending is not about handing out money. I can literally stand on the street and hand out 500 rupee notes if I wanted to do that. Since the product is money, it has the unique characteristic of being something that everybody wants. I don’t have to convince anyone of all of money’s amazing features!
Lending is actually about getting your money back (and making some interest income along the way of course). A business is successful only if it can collect the monies due. It is impossible to understand how a lending business is doing without knowing what its collection performance/ NPAs is like. Yet nobody asks me about the level of arrears in our book.
If we step back and think about it for a moment, this is strange since there is scrutiny of loan performance for banks and NBFCs. Media report these figures extensively and in recent years there has been a lot of focus from the RBI in cleaning the banking system of bad debt.
So why are the rules of the game different for startups? I think it’s fair to assume that most people are rational so let’s examine some reasons (and their validity) for this different framework:
* Grow first, collect later. I think this is probably the most salient explanation. There is something to be said for building distribution and then optimising collections afterwards. We’re in a world where you can’t wait to figure everything out or you’ll get left behind. Capital begets more capital so all aboard the money train! Investors want to see their companies grow to be the leaders and want to see top-line growth. Until they want to see the bottom line.
* Quantum of losses are small in early stage. Even if a fintech has 15% NPAs, if it has done Rs.10cr of loans then the actual money at stake is only Rs. 1.5cr. Proportionately it is high, but from an absolute perspective it’s not fatal. This mentality is dangerous however: the money at risk will grow exponentially in line with business growth. Further, it may be even worse: the actual % is likely to grow as well, since disbursal growth will happen faster than improvements in collection efficacy.
* It’s too early! i.e. there’s not enough seasoning in the loan portfolio to make meaningful statements. This can be addressed to an extent by looking at cohorts by vintage, but even if a business is new, having a high level of arrears should prompt some kind of introspection. Something is clearly not right. It is true though that we need to go through a few loan life cycles to know how a book is really performing. When there is an economic downturn, to borrow a phrase from Warren Buffet, we’ll see who gets caught with their pants down when the tide goes out.
* This is like other tech right? This is pure speculation, but it could be the case that the investor community and media are used to looking at things through the lens of e-commerce where GMV is what has driven outcomes. Value for these investors is being created by growth, boring things like margins are for later stage investors. This clearly isn’t the foremost reason, but may have some bearing on the margin.
Regardless, while fintech may be relatively new, lending is an old business. The rules of the game are quite simple: you have to get your money back or you’re out of business. E-commerce growth in India has been driven by investor funded discounts and incentives to entice users to their platform. The view was that once aboard, repeated use by loyal (or at least sticky) customers will make the upfront investment worthwhile. In the context of lending, instead of a discount, a lender may relax credit standards to drive growth. In those cases, you don’t make money from repeated use, you’re more likely to just suffer losses.
The Art of Durable Lending
A robust credit model. Sourcing customers is not difficult, figuring out which ones to lend to is the trick. Lending is a business where it’s perfectly ok to turn customers away. Besides businesses that focus on exclusivity, and those with physical capacity constraints, I’m actually struggling to think of an example where they turn away people who are willing to pay their price. Lenders need to sort the wheat from the chaff. This is particularly important in the unsecured lending where there is no collateral. With all the talk about machine learning and AI, I think there is an element of the “emperor’s clothes” going on, but over time these tools will surely contribute to smarter models and will make a significant contribution to decisioning.
We should also expect to see more focus on collections going forward. Collections will assume its rightful position at the top table in management meetings. Again, data can be used to optimise collection strategies for different types of borrowers. A fintech’s ability to build smart systems around collections can be a key source of competitive advantage.
All things said, in the short run, the businesses that crack the distribution problem will be deemed to be the early winners. The real successes in the long run however will be those who can nail the collections problem too.
Next time you ask about how a lending business is doing, ask about collections as well.
