Low-income families’ incomes are small, and can be erratic or lumpy as well. We know that there is a demand for more flexible financial services to accommodate those realities. So why are financial services for this market still so rigid?
Flexibility has outsized benefits for low-income borrowers, as Dean Karlan, Sendhil Mullainathan and many others have pointed out. Most importantly, it helps bring a whole new group of borrowers to the table. Formal, rigid contracts limit loan capacity to a borrower’s lower-bound income, not their average income. If I make $100 in a good week but $10 in a bad week, I can’t take a loan based on my average income of $55. I have to take a loan based on my low income of $10 since that’s what the bank is sure I can repay. M-Shwari and other digital credit products have pushed against this constraint by eliminating rigid repayments. But pay-as-you-go (PAYGo) solar providers have gone much farther.
“There is something interesting about [the PAYGo company I use], which is, there is no pressure on the payment…It is to your own advantage to pay, but if you do not pay, it is your own disadvantage.” — PAYGo customer in Cote d’Ivoire
A group of PAYGo solar lenders, notably M-KOPA and Fenix International, have developed a uniquely flexible finance product. A customer pays a deposit and acquires a solar home system that lights up her house. However, she can only use that system on a prepaid basis; she must purchase “days” of energy with her mobile wallet before she uses them. When her days run out, whether she bought 1, 5 or 50, her lights shut off. Assuming a 12-month loan, once she has purchased 365 days, in accordance with her loan agreement, the unit unlocks permanently and becomes her property. This loan leverages two of the key PAYGo attributes: a digital payments channel that permits frequent, small-value remote payments, and a device that can be turned off remotely.
But here’s the kicker: Although the customer has agreed to purchase 365 days of energy, she does not have to pay every day or even exactly within 12 months of acquiring the unit. Most providers welcome pre-payment, and she could receive a “discount” for early completion, useful for farmers with lumpy incomes. But she may also miss a day here or there because of her volatile income, or even a week if there's an expensive family emergency, such as an illness. For those days her house will be dark, but the moment she sends one day’s worth of payments, the lights will turn back on.
In this form of PAYGo lending (there are others), there are no arrears, no negative balance that must be cleared to get back on schedule. This does create uncertainty around when she will finish the contract, which may or may not happen within the contracted 12 months. However, the providers have enough data, covering hundreds of thousands of completed loans, that they can price the loan for the average completion term, thereby allowing fast payers to subsidize slower ones. This does make the loans more expensive, particularly for longer-tenor products, but allows providers to reach more people.
In other PAYGo models, which resemble traditional bank loans, users pay according to fixed monthly schedules and must clear any arrears before they resume payment. But what happens? Customers accumulate months of arrears like stones on their backs, and eventually stop paying altogether. The shift to more flexible models of lending is reminiscent of the move towards pre-paid mobile phone service, which gave customers control over their usage and payments, and enabled widespread uptake of mobile phones in developing markets.
Unlike mobile plans, PAYGo solar flexibility is mostly implicit; the loan agreements may say 12 or 18 months, but customers who fall behind are met with a carrot, not a stick. Companies structure their escalation procedures in different ways to allow a certain amount of flexibility. With most providers, users get SMS reminders when their credit is about to run out and again when it is depleted. From there it varies. In Uganda, Fenix will call a customer, try to understand why they’ve fallen behind, and work with them to structure a realistic payment plan going forward. Other providers begin repossession procedures quite rapidly, at high expense to them.
Some providers go as far as allowing non-payment on a semi-permanent basis, creating a unique type of borrowing. With M-KOPA, if a customer regularly skips one to three days every week, they will never get a phone call, and their unit will never be repossessed. They could pay this way until the loan’s completion, reducing the effective interest rate. These borrowers are taking advantage of another unique facet of flexible PAYGo loans: not only are there no accumulated arrears, there is no compounded interest. The cost of 365 energy days is the same whether they’re purchased over 12, 14 or even 18 months, allowing lower-income households to reduce the monthly price point to a manageable level. Very few households are this extreme; most prefer their lights on full-time. And there are well-defined, pre-set limits to the amount of flexibility that is offered. But this type of borrowing does extend solar access to people who otherwise could not afford it.
Flexibility has costs, of course. PAYGo loans are more expensive than they would be with arrears and compounded interest. Assessing them as one would a traditional lender becomes tricky: determining the quality of a portfolio of loans, for example, requires more nuanced analysis when 30 to 40 percent of the borrowers have “missed” a payment in the last three months. And there is still value to the company in getting paid quicker, so incentives like upgrades and financed appliances are made available to faster payers.
It is not yet clear which of these aspects, if any, can be applied to non-asset finance. But the growth of “smart” devices is coinciding with more innovative consumer financing in developing markets, meaning that remotely secured collateral could become the rule, rather than the exception. This may allow finance to become more flexible in general, bringing financial services and life-changing assets to a wider group of people.
I like how well this piece explains the model and the nuances. Yet I have a very direct comment: out of the whole piece, only 1 paragraph is dedicated to why it is difficult or "may" not work (see below.)
PayGo is certainly an innovative model, and aiming at being disruptive, but do we even know is PayGo will be in business in a couple years?
For me and many businesses in this domain I know, it would be so beneficial to spend a bit more words on why this kind of model is in fact, really hard to go mainstream, and what will inhibit players from scaling it. When, for example like you say "30-40% of borrows miss payments" and you algorithms and business modeling has to account for that, how much less dependable does your modelling become? Likeliness of affecting your business success?
This would be so useful to see written in the context of the model you describe.
I say this because, now in 2017 in digital finance, if we want to really see scale this is where we need to be blogging.
-- "Flexibility has costs, of course. PAYGo loans are more expensive than they would be with arrears and compounded interest. Assessing them as one would a traditional lender becomes tricky: determining the quality of a portfolio of loans, for example, requires more nuanced analysis when 30 to 40 percent of the borrowers have “missed” a payment in the last three months. And there is still value to the company in getting paid quicker, so incentives like upgrades and financed appliances are made available to faster payers. -- "
Thank you Jill for your comment; these are really important questions! To begin with, I agree that this model may be difficult to apply to unsecured, non-asset lending, although we have seen utilities begin to experiment with offering financing where repayments are bundled with traditional energy tariffs, which could be useful in providing more flexibility. But to get to the core of your question: there are a number of things that may prevent PAYGo from going mainstream (working capital, underwriting, mobile payment infrastructure), but I don't believe that flexibility will be one of them. Leading providers have already accumulated masses of repayment and usage data, which can be used to differentiate problematic payment lapses (e.g. 2-3 weeks early in the loan) from normal payment lapses (e.g. 2-3 days halfway through a loan), as well as to segment customers into fast repayers, on-time repayers, slow-repayers, and likely defaults. This level of customer insight allows them to 'portfolio price' the loans for actual repayment behavior, for which they have quite good historical evidence. Another key to this is having a relatively high-margin product; as competition increases in the sector it will be interesting to see if downward pressure on prices also affects flexibility, and which is more important to customers in the long run.
I like the initiative. At least someone takes risks to save the poor. Otherwise everyone would have left them behind with proper business excuses.
A visionary company thinks higher than just making money