Is POS Lending a payments service?

No, these are lending products with incidental payments functions

Key insights in this post

This post uses the availability of Affirm’s data-rich 2024 earnings presentation as a window into the POS Lending market. It tries to determine whether this category is a payments vertical or a lending vertical.

• POS Lending has two distinct products

  • Pay-in-x largely caters to debit-centric consumers with short-term, low ticket lending at 0% APR

  • Digital installments caters to a more credit worthy population, although almost half are non-prime. The loans fund medium ticket purchases

  • In both cases the merchant pays an MDR to the lender that helps fund the loan

• The products primary advantage is being in the digital purchase path

  • Consumers can judge affordability better when they see the cost as “x payments of y dollars each”. In some cases there is 0% APR or subsidized APR

  • Merchants can link usage to a sales lift, either increased conversion or increased basket size. They are willing to pay MDRs higher than debit or credit to capture those higher sales

  • Cards only appear at checkout, after the consumer has already decided to buy and cannot demonstrate a similar sales lift; therefore interchange simply seems like a burden

  • The Point of Sale version of these products are invoked at checkout and also can’t be directly linked to higher conversion or basket size

• Neither product is used for purchasing without borrowing

• As such, these are lending products not payments products and Affirm and its competitors are not Payments Fintechs but Lending Fintechs

Introduction

Affirm published earnings last week and fell just shy of GAAP profitability for the quarter. This is a big improvement, and the trends are even more encouraging. This post will use Affirm’s data-rich presentation as a lens into a misunderstood market.

The category has two distinct product constructs that are often conflated:

• Pay-in-x. These are typically used by debit-centric consumers for small ticket purchases. Klarna & Afterpay are among the pioneers. These loans are only 15% of Affirms portfolio

• Digital installments. These are typically mid-ticket loans and may be issued to consumers with credit cards. Affirm is best known for these. Installments account for the other 85% of Affirm’s portfolio, with a small portion at 0% APR and the bulk Interest Bearing. When Affirm started, a bigger portion of their loans were 0% APR, subsidied by the merchant, especially Pelton; today’s higher interest rates have made 0% APR much rarer

A word on nomenclature. Both constructs are sometimes referred to as “Buy Now, Pay Later” (BNPL). That confuses matters. I typically reserve BNPL for Pay-in-x loans, but the market overall isn’t as scrupulous. So when you read other commentary make sure you understand when they are using it only for Pay-in-x or also for Installments.

All these products started in eCommerce but have migrated to in-person as well. Originally, individual lenders focused on one product, but most have expanded to cover both. Both have high growth rates, although growth has slowed from historical levels. At Affirm, the Pay-in-x category grew 23% YoY while the interest-bearing Installments category grew 33%.

Providers claim two key advantages for these products:

1. Consumers prefer the fixed term model because it keeps them out of the debt trap associated with revolving credit

2. Merchants prefer these products because they demonstrably lift sales. The loans are embedded in the digital purchase path so the linkage to sales is explicit

Putting my cards on the table, I am a fan — but I think these key claims are overstated. The exam question here is: Are these payments products or lending products, and therefore, are the providers Payments Fintechs?

The products

Installment loans to finance discrete consumer purchases are not new. At the high end, an auto or boat loan made in a dealership is a POS loan. So are “sales finance” products available for furniture purchases, appliances, jewelry and other high-ticket consumer goods. And at the very high-end, home improvement loans fit the criteria. Historically, applications used a 3-part carbon paper form and checked credit via phone, email, or fax. The last time I bought furniture, those 3-part forms were still in use.

The Fintech versions are distinguished by digitization. Lending is embedded into an end-to-end digital journey. This confers an advantage over cards since the loan offer is in the purchase path rather than at checkout or afterwards. We have all seen this in online shopping: On the product page, an offer divides the purchase price into “x payments of $y per payment”. In theory, those terms increase conversion and basket size by spreading out the consumer’s cash flow.

In the early days, many of the installment offers had 0% APR subsidized by the merchant, and Pay-in-x still does. Higher interest rates made 0% APR offers impractical on most bigger, longer term loans. At today’s rates, most merchants can’t afford a subsidy down to 0%. In Affirm’s disclosures, the MDR on 0% APR installment loans in 12.5% — 4x even the most expensive credit card. But 0% long-term loans are only 13% of the portfolio. As interest rates decline, these may come back in fashion — Affirm actually reports 70% YoY growth in the 0% APR category.

Pay-in-x

In Pay-in-x loans, “x” is usually 4, but some providers now provide a range of payback periods. Repayments are aligned to consumers’ bi-weekly pay periods. The most common format is 25% paid at purchase time and another 25% every two weeks – so the full loan is paid back within 6 weeks. This spreads payments over 4 pay periods.

These loans are distinguishable from most conventional lending:

• They cater to a debit-centric audience. A credit card user gets no cash flow benefit from Pay-in-x because 2 of the four installments happen earlier than they would on a credit card and only 1 payment occurs later than a typical credit card due date

• They never have an APR. The merchant pays an MDR to the lender to cover the interest and processing costs. Affirm discloses this MDR as averaging ~5% — this is higher than even the most expensive credit cards and much higher than regulated debit at ~25¢ per transaction. To pay this much, the merchant has to believe that the consumer would not do the transaction otherwise, or would have spent less

• Payback is typically via debit card. A Pay-in-x loan creates 4 debit transactions to complete one purchase. Durbin-regulated DDA banks should love this product! They wouldn’t have made the loan anyway and it quadruples their debit income

According to surveys, 90%+ of Pay-in-x users are debit centric, but some credit card holders do use it. When I looked at the characteristics of these users, they are almost always revolvers using Pay-in-x to avoid adding to their balance. They are effectively using zero-APR Pay-in-x to avoid borrowing more on their high-APR credit card. As much as debit banks should love Pay-in-x, credit card issuers should hate it.

Pay-in-x competes with another high-growth innovation — Earned Wage Access (EWA). In EWA, the consumer borrows against their accrued wages. This is also interest free, but the consumer often pays a fixed usage fee. Sometimes the employer subsidizes this. Pay-in-x & EWA are tapping the same need for small dollar, short-term credit. The other alternatives are high-APR products like Pay Day loans, Pawn Shops, and Title Loans. Some banks now offer “advances” that also fit this need, but these incur a fee.

To me, Pay-in-x is a miracle product – it provides a little bit of credit, at low/no cost, for a short period. This is what activists have been wanting for decades — to avoid “predatory” alternatives. But, Pay-in-x draws criticism anyway. The main critique seems to be that if used too often it can become a debt trap that drives credit scores even lower. But for someone with no credit, even this outcome is no worse than where they already are. I don’t get it.

Returning to our original question: Is Pay-in-x a payment product or a lending product? My conclusion is: It is a lending product. There is no reason to use it except for the extended repayment terms. As a payment method, it has more friction than basic debit and it has no other consumer benefit beyond extended payback.

Digital installment loans

These are Sales Finance loans updated for the digital age. They are embedded in the digital purchase path, well before checkout, and are underwritten in real time. They are largely used for mid-ticket purchases.

The original success story was Peloton. This was an expensive purchase that Affirm financed. Peleton subsidized the APR down to zero as market interest rates were low anyway and the margin on the bikes was high enough to afford the MDR.

0% APR offers were almost too good to refuse. Even cardholders with the ability to purchase in a single payment, might extend terms at no cost. As the loans became more popular, versions without 0% APR were introduced, for less credit-worthy customers. As interest rates rose, this became the norm. Today, 72% of Affirm’s loans carry a card-like APR yet the product is still growing fast.

This is likely due to positioning in the digital purchase path. A consumer is exposed to these products as they shop, not just at checkout. That allows the consumer to gauge affordability and buy a bigger basket. Merchants pay ~2.5% MDR for that improved conversion. Note that at small merchants, this is roughly equivalent to what they pay their acquirers. At large merchants, on IC+ pricing, this is equivalent to high-end credit cards, but for lower FICO customers.

It is an open question whether the transparency Affirm rightly brags about has anything to do with popularity. Affirm charges simple interest, has no late fees, and avoids other hidden fees. All of this certainly accrues value to the Affirm brand, but my suspicion is that position in the purchase path matters much more.

Affirm also claims it can underwrite more precisely because it knows the SKU of what is bought. I thinks this help detect fraud rather than fine-tune credit risk. Many years ago, I worked with a European fraud vendor who told me a story on this topic. A high-end linen manufacturer ran a promotion on “crib sets” for newborns. With each purchase, the buyer got a free digital camera, back before smart phones made those obsolete. The promotion was very popular, but for the wrong reason: A huge percentage of the sales were fraud — crooks were buying the bundle, throwing away the linens and fencing the digital camera. The promotion was short-lived.

Many lenders claims that younger generations prefer installment loans over revolving credit – to avoid a debt trap. I suspect this is an artifact of the credit crisis. At that time, the card issuers raised credit standards and reduced lines, with the main impact falling on new-to-credit consumers. That segment could get an online installment loan when they either couldn’t get a credit card or they couldn’t get a card with enough line. I have seen no evidence of younger consumers eschewing cards today.

In response to this competition, the big credit card issuers will now convert a purchase into a fixed-term tranche within a card credit line. I have seen no market data on how popular this is, but if the “Gen n prefers installment loans” hypothesis were true, these variations would be popular enough to brag about and disclose.

Of course, when digital installment products were new, many came with 0% APR, which was enough incentive on its own. So, if younger, creditworthy consumers ever did prefer installments I suspect the preference has faded. Maybe not in surveys, but in actual behavior.

These loans compete with payments products like PLCC cards, Cobrand cards, and issuer-branded credit cards. Debit-centric consumers generally don’t qualify. PLCC and Cobrands often subsidize interest as well or offer big discounts for enrolling (“15% off your first purchase!”). The PLCC merchant builds a database of loyal customers. Regular credit cards pay rewards, so unless you need to borrow, you are better off spending on such a card. While all these products co-exist at many retailers, My observation is that POS Loans lead at merchants with low repeat business while Cobrand/PLCC products lead at merchants with many repeat visits.

All these products lift sales in theory, POS Loans can better prove conversion but Cobrand/PLCC can better prove merchant loyalty. Cobrand cards contribute revenue share above their loyalty value as well.

Affirm claims to charge a ~2.5% MDR on interest-bearing loans but discloses that 43% of receivables are “non-prime” which they define as FICO scores below 660. So the merchant is paying an MDR as if all the borrowers are super prime when in fact most of them are not. Merchants need to believe that being in the purchase path improves conversion or increases basket size to pay such a high MDR. Given the borrowers are typically credit-constrained, this is likely true.

If we go back to our exam question, Digital Installments are a lending product, not a payments product. Yes, they convert a sale, but the consumer only uses them to borrow – never just to spend. The question was more debatable in the 0% APR era, but now that most loans carry an APR, that debate is over.

Extension to in-person commerce

Both Pay-in-x & Digital Installments have added Digital Wallets as a distribution channel. This is mostly to extend from eCommerce to in-person purchases.

The POS lenders needed a way to expand from eCommerce to the larger in-person market. They have done so via NFC wallets and virtual cards. The consumer can use an App to request a loan and, if approved, a virtual card is pushed into an NFC wallet (e.g., Apple Pay). Since the virtual card carries a 2%+ interchange rate it substitutes for the MDR the lender charges an online merchant. This model is more useful for Pay-in-x than for Installments. But, it is not in the purchase path and is clunkier than just paying by card or standard Apple Pay. You have to want to borrow to use it.

Apple Pay originally offered these products directly, but has since outsourced to Affirm. Citi just announced that its “Flex” service will be available through Apple Pay; Flex allows a cardholder to structure their purchase as installments after purchase.

The dependence on virtual cards is a potential flaw as these may be steered against if the pending merchant settlement limits the networks' "honor all cards" rule. For Pay-in-x, the merchant would be paying a 2%+ interchange rate instead of much lower debit interchange. Online, they can negotiate the MDR and choose not to offer the product if they don’t think it lifts sales. But, today, the merchant has no choice at the POS even though the lender can’t really demonstrate sales lift.

Interestingly, Affirm just announced a partnership with FIS to add Pay-in-x functionality to standard, bank-issued debit cards. FIS largely serves so-called “exempt” banks; banks with <$10B in assets are exempt from the Durbin cap on debit interchange. Uncapped debit can range from 80bp to 150bp depending on merchant vertical

• This generates less revenue for Affirm than the virtual card model, but perhaps has lower fraud risk, credit write-offs, & CAC to balance the scales

• The banks get the extra payback transactions instead of just the purchase transaction. A portion of debit interchange is fixed, so this raises bank revenue. For example, the Exempt rate for card-present Retail is 80bp + 15¢. In a Pay-in-4 model, the 80bp generates the same revenue, but the bank gets 45¢ extra for the 3 additional payback transactions. Neat trick!

Conclusion

Digital Wallet distribution does not change the fundamentals. These are lenders not payments processors. They have high growth because they are lending to under-served populations; as such, they are not attriting much volume from incumbent card issuers. The Pay-in-x product mostly serves debit-centric consumers and the Digital Installments product has a big non-prime user base.

That means Affirm and its competitors are not Payments Fintechs:

• Pay-in-x users all have a debit card but generally don’t qualify for affordable credit elsewhere. Pay-in-x spreads what would be one debit purchase over several debit transactions, but the payments are all on debit. The product has no value except to extend payback at 0% APR – but it is very valuable for doing that

• Digital Installments are the closer call. For some customers, they compete with credit cards for revolve, but mostly in lower FICO bands. Without the 0% APR option, credit card holders don’t really need to use this, particularly if their card issuer can also break a purchase into installments

That analysis establishes that POS Lenders are not Payments Fintechs at all; yet they are often grouped in with that segment – including by me. Why? They do help complete a purchase, but their reason for being is to finance that purchase not to execute it. Clearing and settlement is just a necessary evil here, not the key deliverable. And these lenders are more exposed to credit risk, fraud risk and interest rate risk than a pure payments company.

The real magic of these products is on the merchant side. It is unsurprising that credit constrained customers will use them, particularly if they are conveniently in the purchase path. What is surprising is that merchants, who complain constantly about high interchange, are willing to pay even higher MDRs for credit-challenged transactions. According to Affirm disclosures, Pay-in-x merchants pay 5% MDR instead of regulated debit interchange at ~25¢. In Digital Installments, merchants pay ~2.5% MDR where credit card interchange is around 1.5% for low-end cards and 2.5% for affluent cards. Merchants pay an affluent rate for non-affluent customers. Strangely, merchants don’t mind paying up if they think they are getting a sales lift.

Going forward, I will treat this category as a distinct Fintech vertical, but not a payments vertical. I still think these are great products, but they aren’t really payments products.