Buy-now-pay-later (BNPL) bellwether Affirm released its calendar 2022 Q4 earnings this past Wednesday, for the period ending December 31st, 2022. Despite the bottom line miss and announcement to layoff 19% of its workforce, the Paypal Mafia / Max Levchin-run credit platform, at least for those who seek to better understand the consequences of the Fed’s higher interest rate regime on consumer credit (in contrast to CNBC’s level-one thinking, bobo “fintech reporters” ), laid out an impeccably detailed analysis of the consequences of higher borrowing costs to Affirm’s BNPL model. In applying the higher interest rate etiological dynamics to Affirm’s business, Levchin, and CFO, Michael Linford, offered extremely valuable insights on how the effects of this particular macro-economic headwind of expensive capital is manifesting itself in Affirm’s ecosystem, with its merchant partners, capital providers, and consumers. Further, and albeit unintentionally, the earnings presentation information and management commentary serves as a fascinating case study for financiers and capital market professionals alike who seek to enhance their predictive wherewithal relating to the same, in its application to the private credit markets and consumer financing landscape more broadly.
Part 1 – The Model
The financing model that Levchin assembled to fund and originate loans for Affirm’s business is highly sophisticated. It’s best understood if broken down into its two core pieces: the consumer financing side and the capital provider side.
On the consumer side, the crux of Affirm’s BNPL market appeal stems from the transparency of its credit terms, the ease of access and application at the point of sale, and the unique decisioning algorithms that approve consumers, not for blanket credit lines, but solely for the credit needed to purchase a specific product or service. But the most powerful driver of consumer appeal and demand is the financial engineering that permits the 0% financing costs for short term (4 to 12 weeks) installment plans on relatively small ticket sizes. Accepting that there’s no such thing as free lunch, the 0% financing is only true to the consumer. In reality, the financing costs for those installment loans are borne by the merchants (and sometimes the product manufacturers) who admit to a haircut on the price that the product or service is sold for. For example, when a consumer opts to purchase a $100 pair of jeans form a merchant, the merchant is actually agreeing to sell the jeans for only $95, ergo the haircut. Affirm then funds the merchant the $95 and goes on to collect the $25 per week in four weekly installments ($100) from the consumer. Thus, the merchant’s 5% haircut turns into a 5.2% return to Affirm on the loan.
On the capital provider side, Affirm is originating microloans which are funded through three primary channels: warehouse credit facilities that allow Affirm to borrow against loans on its balance sheet (balance sheet assets), forward flow relationships with ABS sales to to third party investors, and/or securitization vehicles where loans are packaged-up and slice-and-diced into structured debt (bonds) that are then sold off to investors. This process is very similar to the packaging and sale of mortgage backed securities (without the GSE backstop). Affirm uses these three methods to fund the loans it originates, and if all goes well (investors get the risk/return they’re seeking) ,continue to increase the volume of capital available to it to originate even more.
Part II – The Question
So here’s the big question: Affirm went public in January of 2021 when the Fed Funds rate was 9 bps. The Fed Funds Rate for December 2022, the last month of the just-reported quarter, came in at roughly 400 bps higher than that. How did the rising cost of capital affect Affirm’s business on both the capital provider and consumer side (including the effects on the merchants who subsidize the 0% installment loans)?
To start to answer this question, we need to understand a bit more nuance about Affirm’s business model. Not all of its consumer microloans are 0%. In fact, plenty of Affirm’s installment loans, especially to consumers who make larger ticket purchases over longer terms, require the consumer to pick-up a portion of the debt service. This is purely a function of pricing in the greater risk of the loan. Even though a good portion of the installment loan is still subsidized by the merchant, the larger loan sizes and longer terms result in a risk premium (still clearly stated to the consumer of course) which the merchant either can’t subsidize at all (because covenants in its incorporation preclude it from taking on balance sheet risk with certain risk profiles), or can’t subsidize entirely (the economics make it simply too expensive).
Part III – Answering the Question
When the The Fed raises interest rates, that expense directly flows to the private lending/credit markets that provide the capital to fund Affirm’s originations. The private capital markets – private credit providers and secondary market debt purchasers – have to pass along the higher cost of capital to Affirm, which in turn has to pass along the higher cost of capital to its participating merchants and consumers. This must happen to protect the respective parties’ spreads. Otherwise, the lenders – all of them in Affirm’s financial ecosystem – can’t generate positive returns and the whole system grinds to a halt.
And this reality feeds directly into answering the key question of how rising rates have impacted Affirm’s consumer credit business.
Very early on in the earnings call, two crucial, yet seemingly incompatible (I’ll explain shortly) statements of fact were disclosed:
#1 for the quarter, Affirm saw its loan book composition shift away from 0% loans toward an increasing percentage of interest-bearing loans, and
#2 loan delinquencies fell on a sequential basis
That loan delinquencies fell as interest-bearing loans increased seems, well, backasswards. Traditional thought predicts that larger, longer-term loans carry greater attendant risk, and by extension, higher chances of delinquency and/or default.
But that’s exactly not what Affirm is seeing in its business model. In fact, according to management, it’s seeing loan performance improve. This is incredibly important to grasp because in tighter credit cycles with higher interest rates, non-performing loans tend to increase because they cost more to the borrower in repayment. They “squeeze” the consumer. The conclusion to be drawn here is that higher interest rates are not negatively affecting Affirm’s business with higher delinquencies.
Part IV – Answering the Question, But Still Have No Clue How.
In order to square this circle and get to the non-intuitive answer of how higher interest rates are affecting Affirm’s business, there’s only one more place to look, and that’s back at bullet #1 above.
Higher interest rates are not squeezing Affirm’s active users (yet!), they’re squeezing the merchants who are subsidizing the 0% loans. This is evidenced by the growing proportion of interest-bearing loans in Affirm’s loan book. Merchants can only take so much of a haircut to subsidize these loans: they can sell a $100 pair of jeans for $95, but they can’t sell it for $88. Therefore, there’s no other option than to pass through the increased borrowing costs to the consumer, who is now seeing less options for 0% deals, and necessarily, more options for interest-bearing options, causing the change in loan book composition that Affirm management is seeing.
With this shift to more interest-bearing loans in Affirm’s loan book, the capital providers and investors (and Affirm, so much that it too holds these loans on its books) now have to adjust their risk scoring models to account for the change in loan characteristics they’re purchasing (or in Affirms Case, selling) in order to properly re-price (higher) the capital allocated to capitalize Affirm’s origination machine.
Part V – Answering the Question. What the Hell is Going On Here!?
Affirm mentions this in the earnings release by acknowledging that it needs to reprice its offerings with merchants. However, even so, this doesn’t seem to be affecting demand for Affirm’s products, nor appear to be a threat to its business model. By all accounts Affirm is managing (very well, in fact) to control the effects of rising rates on its business by limiting the increase of borrowing costs passed-through to the consumer. The continuing demand is further supported by the numbers: continued growth in GMV, active users, and active user transactions.
So rising interest rates are affecting Affirm’s business by decreasing the availability of 0% loans and passing along greater borrowing costs to consumers in the form of interest-bearing originations. But, yet, by all accounts, as represented by its key performance indicators, the higher interest loans are having a net positive effect because the business is still growing and consumer demand for its products is still increasing???
The answer, or the riddle wrapped in a mystery inside an enigma, as to how can this possibly be, is directly tied to Affirm’s core technology: underwriting and credit scoring. As long as delinquencies don’t accelerate (which they haven’t, they’ve actually slowed down), this constitutes hard evidence that Max Levchin has indeed developed a superior credit decisioning technology. And this superior technology, the underpinning of the platform, by keeping delinquencies and non-performing loans in check, is acting as a powerful countervailing force against the deleterious effects of rising interest rates on its core consumer finance product.
Flying in the face of traditional finance and macroeconomic orthodoxy, this true fintech offers a case study sporting a poignant counter-example to conventional predictive models, by keeping the pernicious effects of rising interest rates on Affirm’s consumer credit business from meaningfully impacting demand because Affirm’s superior underwriting technology is keeping delinquencies down and preventing its cost of capital, not from going up, but, from going up too much!