For three days each week during the month of April in 2014, a seasoned product manager named Lulu Young, an engineering manager named Paul Connolly, and a 24-year-old jewelry salesman named Nick Molnar gathered in a bare, windowless conference room in Melbourne to hash out the features and functionality of a financial product that existed only in Molnar’s head. The goal was to appeal to two constituencies at once: online retailers, who were always eager to convert more virtual browsers into actual shoppers; and consumers, some of whom didn’t have credit cards but, Molnar thought, might still like a way to get their goods first and then pay for them over time.
The product that emerged gave would-be users two options, both of them interest-free, which placed them outside Australia’s credit regulations. The first was called Pay After Delivery, which let people wait 30 days before handing over any money, not unlike using a credit card. The second option, Pay Over Time, let people split their bill into four installments. It didn’t take long for Young and her colleagues to agree on the structural details: They’d need to let people stretch payments over more than 30 days, but “we didn’t think that a much longer period was necessary and looked at 60 days,” she says. “Two payments made it too similar to the existing paradigm: payments roughly every month. The mental gymnastics for four was straightforward—halve any number and halve that again.”
They called the service Afterpay. Later that year, shoppers at the jewelry website Molnar had started with his brother and their jeweler parents were the first to encounter the two “new” ways to pay. A hoped-for flood registered as little more than a trickle, and it didn’t take long to see that the Pay After Delivery option wasn’t resonating much at all. The mood shifted when Princess Polly—a website that today bills itself as “the ultimate global fashion destination for trendsetters who want the latest Insta-ready, TikTok-approved, celeb-worthy looks” and whose young female clientele didn’t include many holders of credit cards—experimented with the pay-in-four option in May 2015. The system sent Young an email each time someone chose to pay in four. “The trickle of orders became a steady stream almost overnight,” she says. “I quickly got rid of the alert.”
Pay-in-four is now ubiquitous in Australia and fast approaching that status in the US and Europe, accessible at the checkouts of hundreds of thousands of retailers online and off. It powered the ascendance of Afterpay, which attracted investment from Chinese tech giant Tencent Holdings Ltd. and the US investment firm Tiger Global Management LLC and made Molnar and his co-founder, Anthony Eisen, billionaires. (Digital payment company Block Inc. bought the company in January for $29 billion, making it the most valuable acquisition in Australian history.) The model has become the flagship product of “buy now, pay later” companies. Lauded as a much-needed alternative (and threat) to credit cards and predatory lenders and criticized as a gateway drug to debt for the young and inexperienced, BNPL represents one of the biggest and fastest changes to consumer credit in decades. In the US it took root in 2018 as a way to buy clothing, cosmetics, and other discretionary items and exploded in popularity amid the pandemic. You can buy now and pay later for just about anything, including aspirational big-ticket items, like Pelotons and designer couches; trifling and small things, like socks and underwear; and dire necessities, like groceries and gas. There are even small BNPL companies that will help you buy a gun.
Companies such as Afterpay and its main competitors, Klarna and Affirm, often attribute their phenomenal growth to widespread distrust and dislike of credit cards, particularly among the under-40 set. Never mind that this doesn’t comport with the data (a September study by TransUnion showed that BNPL users possess more credit cards than the general “credit-active” population), the narrative has been repeated so many times it’s become practically gospel. But a deeper look into the pay-in-four model suggests some less facile explanations.
The first relates to a kind of regulatory arbitrage, in which pay-in-four products bypass consumer protections designed to prevent people from getting screwed or screwing themselves. Particularly in the US, BNPL companies haven’t been subject to much of the regulatory oversight normally applicable to entities extending credit. Take the Truth in Lending Act, a landmark law enacted in 1968 and amended many times since, which requires extensive disclosures for unsecured consumer loans split into payments of five or more. That law doesn’t touch pay-in-four.
The second explanation has to do with how pay-in-four taps into human psychology and consumer behavior. It satisfies what behavioraleconomists call our present bias, which drives the desire for instant gratification. It also exploits the fact that unexpected losses almost always exceed unexpected gains: Humans tend to discount the probability of bad things happening in the future (we’ll lose our job or the car will need repair, making it hard to repay) and overestimate the probability of good things to come (we’ll be able to work overtime next week so it’ll be easy to repay). And because we are much less sensitive to the pain of a price tag if individual chunks feel affordable (parting with, say, 25 of our hard-earned dollars today, then three more times in the near future, hurts less than parting with $100 all at once) we are more willing to spend more in the aggregate. That in turn explains why merchants love pay-in-four. But even as the industry has grown spectacularly—the number of people who’ve tried the services in the U.S. has climbed 300% every year since 2018—nearly everything about the business environment has recently turned more serious and challenging.
Shop online, and you’re bound to encounter offers from Afterpay, Klarna, or Affirm to let you split purchases up to $2,500 into four installments. You enter the usual information, concluding with a credit card, debit card, or bank account number to pay one-quarter of the total price upfront. The companies then use proprietary predictive models to assess the riskiness of lending you money to complete the transaction. Requests get approved within seconds. An email confirmation invites you to download the company’s phone app. The app allows you to track the remaining three biweekly payments, which can be automatically charged to the card or account entered at checkout. It also functions as something more clever: an online shopping mall, where you can browse the wares of brands that have formed partnerships to accept that BNPL provider as a form of payment. Copious eyeballs, dollars, and data flow as a result.
The arrangement is like layaway, but in reverse. Make your payments on time, and the pay-in-four model is all upside: You’ve borrowed money free of any interest. But fall behind on your payments, and you might get hit with late fees from the BNPL provider. If your account balance is running low, the auto-deduction might trigger an overdraft fee from your bank. If you use a credit card and don’t pay off the balance in full, you’ll end up paying interest indirectly. BNPL companies slam the evils of credit cards but accept them as a way for users to pay on their platforms.
Merchants benefit from accepting credit cards—and paying a transaction fee averaging 2%—because doing so increases sales. BNPL one-ups the incentive: Merchants pay an even higher fee—up to 6% for pay-in-four—but are rewarded with more completed transactions (less “cart abandonment,” in industry parlance) and even higher sales. These fees from merchants make up the single biggest source of revenue for many BNPL companies, a fact they cite as proof that their interests, unlike the interests of credit card issuers, align with those of customers. Card issuers, by contrast, earn the vast majority of their revenue from fees and interest heaped onto users who don’t pay off balances in full and on time.
There is research showing a correlation between boredom, browsing the internet on your phone while lying in bed, and spontaneous shopping sprees—so maybe it was no wonder that the BNPL industry took off as the pandemic dragged on, making hay and minting billionaires. Consumer advocates watched with concern: What exactly was BNPL, anyway? The companies operating under the moniker offered significantly different services. Afterpay stuck to interest-free pay-in-four and, like Klarna, charged late fees. Affirm and Klarna did pay-in-four but also issued interest-bearing long-term installment loans. The lack of uniformity could be confusing. And it wasn’t clear what laws or consumer protections applied. Congress had passed the Credit Card Accountability Responsibility and Disclosure Act of 2009 to make it harder for credit card companies to market to young people and risk saddling them with debt, and here were BNPL companies, unbeholden to the law and appearing to do just that.
Was pay-in-four even credit? New services can be hard to categorize, and, just as deferred-presentment providers (aka payday lenders) did back in the 1990s, BNPL companies took semantic liberties. Afterpay referred to itself as a “budgeting tool.” Klarna called itself a “global payments and shopping service.” As time passed, people called their bluff. “If I spend your money now, and then I pay you back with my money later, common sense would dictate that I am borrowing from you and that you are lending to me,” Ritchie Torres, a Democrat representing the South Bronx who’s on the House Committee on Financial Services, said at a November hearing on BNPL. Regulators in California and Massachusetts have accused BNPL providers of offering illegal loans and forced them to register as licensed lenders.
The Consumer Financial Protection Bureau took the first step toward bringing BNPL products under its regulatory umbrella in December, when it ordered companies to submit information about industry practices and risks. The CFPB’s very creation was predicated on the need to oversee the surging, loophole-enabled popularity of innovations like BNPL loans, says Robert Lawless, a professor and expert on bankruptcy and consumer finance at the University of Illinois’s College of Law. “Financial regulation is a game of whack-a-mole,” Lawless says. “There’s always going to be some new device or transaction that the existing statutes don’t cover, so we need a regulatory agency to track these things, because Congress—even a functioning one—can’t keep up.”
Why didn’t the CFPB act sooner? The rise of BNPL coincided with the era of Donald Trump, whose administration was explicitly hostile to the agency and its mission. As a congressman, Mick Mulvaney, Trump’s first CFPB acting director, had sponsored legislation to abolish the agency, which he called “a joke” in a “sick, sad way.” His successor, Kathy Kraninger, who had no prior financial regulation or consumer protection experience, tried during her tenure to make it easier for payday lenders to loan to people who couldn’t afford to repay. The lag may have also been driven by data: The CFPB uses consumer complaints to help guide where to use its limited time and resources, and the number of BNPL-related complaints didn’t explode until 2021. This year complaints are on pace to blow past last year’s total of 547. The most common ones concern incorrect information on credit reports and attempts to collect debt that isn’t owed.
In an interview, Rohit Chopra, appointed by Joe Biden to lead the CFPB, is clear-eyed about the benefits and risks of BNPL. Enabling people to overextend themselves, he says, has “led many people to believe that what they thought was affordable or what they thought was free was not.”
Chopra issued the request for information on BNPL companies two months after assuming office and says he expects the agency to issue its initial report on the industry this fall. The CFPB could eventually require companies to abide by consumer protection laws that don’t now apply. It could subject the biggest players to regular examinations. Beyond rulemaking, which can take years, the bureau can issue best practices and fines for noncompliance. The industry may complain that this is regulation by enforcement, but it’s where the CFPB can move the fastest.
Opinion is divided on whether increased regulation could curb the BNPL industry’s growth. Juniper Research suggests that regulatory changes will merely place it “on a more secure footing.” Investors, however, have grown increasingly worried about the viability of companies reliant on lending to mostly younger, often subprime consumers, particularly if there’s a recession. The shares of Affirm and Block, Afterpay’s parent, have lost more than 55% of their value since the start of the year. That’s made it harder for Klarna, which is private, to raise capital: Its July funding round slashed its value to just under $7 billion, from $46 billion in mid-2021.
Losses, meanwhile, have grown. In the first three months of the year, Block’s more than quadrupled compared to the same period a year earlier, to $91 million, driven primarily by delinquency on Afterpay loans. In that same quarter, Klarna said credit losses jumped 51%, to 1.19 billion Swedish krona ($115 million) because of its expansion in the US, France, and elsewhere. At Affirm, charge-offs jumped 362%, to $67.2 million. (Lenders charge off a loan they’ve decided a consumer isn’t going to repay, refer it to a debt collector, and record it as a loss on their books.)
Competition has also increased, with some two dozen BNPL-branded companies now operating in the US, according to the Federal Reserve Bank of Kansas City. In June, Apple Inc. announced it will introduce a pay-in-four option called Apple Pay Later. PayPal Holdings Inc. inaugurated its version in August 2020. A race to the bottom has begun, as companies offer ever-lower merchant fees to gain market share. “Lenders will need to find other sources of revenue to maintain growth and profitability,” the CFPB said in December.
Because it’s new, BNPL hasn’t been tested during a meaningful economic downturn. With inflation high, money doesn’t go as far—which suggests demand for BNPL could rise, as it did for traditional layaway during the Great Recession. But if people curb spending altogether, usage could dip. The number of people unable to repay their loans could climb. The industry is in the early stages of a defining test.
In early 2018, four years after its initial crew gathered in Melbourne, Afterpay exported its pay-in-four model to the US. Klarna and Affirm already existed, but neither offered a pay-in-four service yet. Klarna, which began as a payments processor in Stockholm in 2005, is now the biggest BNPL provider by sales volume, according to Bloomberg Intelligence, with 400,000 merchant partners and 150 million active users in 20 countries.
Affirm, based in San Francisco, started out 10 years ago offering long-term installment loans. It earns a significant chunk of revenue from interest on them, albeit in a way it bills as more transparent than a credit card. Affirm doesn’t charge late fees, but it cuts you off from the platform if you don’t pay. Max Levchin, the chief executive officer, founded the company with Palantir Technologies Inc. co-founder Nathan Gettings and two others after co-founding PayPal with Peter Thiel and Elon Musk. Levchin is particularly outspoken about what he sees as the potential for pay-in-four and other BNPL installment plans to usher in a new era of consumer finance. They’re tools in the service of his bigger vision: “reinventing lending” by “revolutionizing” the credit scoring system.
It’s hard to be a fully functioning member of the economy, and thus society, without access to credit. For decades, credit reporting companies Experian, Equifax, and TransUnion have functioned as gatekeepers—central repositories for data about the income, assets, payment history, and outstanding debt of American consumers. When you apply for a credit card or a loan, lenders use that data to help determine your creditworthiness, or how big a risk you pose to lend money to.
This traditional system has long been reviled. Data show almost 106 million Americans are considered “credit invisible,” “unscorable,” or subprime and below. Poor people, people of color, immigrants, and young people disproportionately fall into those categories and get cut off from traditional banking and the advantages it can bring. Black and Latino people tend to have lower credit scores. Therein lies the potential for fintech innovation to mitigate racial, financial, health, and wealth gaps, Kristen Broady, at the time a fellow at the Brookings Institution, said at the BNPL hearing in November. “Through technology and automation, they can reduce costs and prices, speed up delivery, and increase convenience for underserved populations,” she said.
BNPL companies speak in the language we’ve come to expect from Silicon Valley: The legacy credit system is broken and unfair; what they’re creating to replace it will make the world a better place. Speaking from his home office, Levchin describes his inspiration to create an alternative. He came to the US from Kyiv as a teenager and tanked his credit score in his penniless youth. It took years for it to rebound. “This is a personal conviction based on direct experience,” he says. “The system was ripe for fixing when I experienced my road bumps all those years ago, and I don’t think it’s changed all that much.”
That’s where the algorithms come in. Instead of running a one-time assessment on the creditworthiness of an individual borrower, as credit card issuers do before they issue a card, BNPL companies assess the risk of each individual transaction. Making every transaction “an explicit borrowing event,” Levchin has said, protects companies from excessive risk and borrowers from overextending themselves. And it qualifies a lot more people for loans.
Take a $5,000 line of credit: A traditional lender must factor in a borrower’s income, expenses, and obligations. It uses credit scores to predict the likelihood that the borrower will pay 90 days late (or more) over the next 24 months. Contrast that with a $500 loan for headphones. The BNPL provider receives a relatively hefty 25% down payment (the first of the four payments) upfront, and need only estimate the likelihood of default on the remaining balance over the next six weeks. The smaller-dollar nature and shorter repayment window add up to lower risk.
The precise components fed into the algorithms aren’t disclosed. Afterpay’s models include “hundreds of internal and external data variables and features,” says spokeswoman Amanda Pires, declining to give more details. Affirm’s “proprietary technology that we’ve been developing for a decade” uses “proprietary data we’ve been building up for nearly as long,” Levchin has said. “Our process involves looking at credit report data, but could also involve some Affirm-specific stuff, like what we know about the merchant and the thing they are about to sell you.”
The resulting loans get branded by BNPL companies as tools of empowerment and freedom. Users skew young—so less financially experienced and savvy, by definition—and female, many with low to moderate incomes. BNPL, the companies say, transforms them into responsible customers shielded from predatory lenders—and advances, loan by loan, the cause of financial inclusion.
The industry has faced accusations that it emotionally manipulates users and glamorizes debt, using social media influencers to drive adoption and marketing slogans such as this one from 2018: “Broke AF but strongly support treating yourself? Afterpay is now instore.” Using pay-in-four for groceries and other everyday essentials suggests economic precariousness, says Andrew Kushner, policy counsel at the Center for Responsible Lending in Oakland, Calif. That, he says, reinforces the need for stronger protections, “so it’s not going to cause more harm down the line.”
The companies say their built-in protections enforce responsibility. They do such things as cap late fees at 25% of the purchase price or start you with a low ceiling of $100 before increasing your limit after you’ve demonstrated your ability to pay on time. But plenty of people truly don’t have the wherewithal to take on debt, and BNPL offers the chance to do it anyway. The companies have been sued by consumers who got hit with bank overdraft fees they didn’t expect. A recent study by economist Amy Crews Cutts showed 17% of BNPL borrowers used the service after maxing out their credit card, and almost half made purchases they acknowledged they couldn’t otherwise afford. In a survey of more than 1,000 users, one-third told Credit Karma they’d fallen behind on payments.
Whether disdain for credit card issuers runs as high as the BNPL industry claims, consumers have developed certain expectations about paying with credit. One is that they will boost their credit score by paying bills on time. Historically, though, BNPL providers haven’t consistently reported the existence of a pay-in-four loan or the borrower’s payment history to the credit reporting companies. That will change; BNPL providers are working with the credit reporting companies to develop a system to do this. But as the CFPB has warned, this lack of reporting has prevented lenders of all sorts from seeing how much a prospective borrower may already owe others. That can result in a practice called loan stacking—borrowers carrying loans with multiple providers at a time. It’s the result of a system that, if not set up to create problems, doesn’t do enough to prevent them.
Chemareéa Biggs, who moved to New York last summer to make it as an artist, is living with the fallout from her own loan stacking. After graduating from the University of Texas at Austin, Biggs rented an apartment in Brooklyn, got a job at a museum, and took on freelance graphic design and teaching work. She’s since lost multiple jobs because of the pandemic, fallen into a debt trap she doesn’t know how to climb out of, had her checking account closed because of repeated overdrafts, and seen her credit score drop by some 200 points. She traces her troubles to a pair of plane tickets, totaling $607, she purchased on pay-in-four plans through Affirm in October and November 2021. “I missed one payment,” she says, “and that’s when things started to go to shit.”
Biggs had found the offer through Alternative Airlines, a travel website advertising various BNPL options. She then took on a series of loans from a more obscure BNPL provider called Uplift, as well as nontraditional online lenders such as Spotloan, which charges interest rates as high as 490%, to cover rent. She now owes a total of $2,588 to four companies. That doesn’t include the $455 Affirm and Uplift have charged off as losses. The experience has revealed how hard it is to keep track of a crazy quilt of multiple loans coming due at different times, Biggs says, and how in the dark she was about the potential risks. “The advertising on these things,” she says, “it almost makes spending look like a game.”
BNPL companies don’t force you to buy things, of course, but they do lower the barriers. A survey by the Federal Reserve Bank of Philadelphia found that contrary to popular belief, convenience, not lack of credit access, was the primary reason people chose BNPL as a payment option. Sebastian Siemiatkowski, co-founder and CEO of Klarna, told Sifted, an offshoot publication of the Financial Times, that he’s questioned internally whether BNPL loans make borrowing too easy. “When it comes to money, friction isn’t a bad thing,” says Alistair Newton, a banking analyst at Gartner Inc. in London. “Sometimes a bit of friction in a payment is good.”
Sometimes there’s friction between customers and merchants. The Fair Credit Billing Act guarantees credit card users the right to file complaints about overpayments and billing disputes and requires card companies to investigate and refund any unjustified charges. There is no such framework for BNPL. Because the companies depend so heavily on revenue from merchants, there may be a financial incentive to keep them happy at the expense of customers.
Thomas Leavitt, a 58-year-old retired bartender, learned that the hard way. In 2018 he shopped online for a ring for his wife. He found one he liked from a Los Angeles jeweler, and when he put it in his cart an offer appeared from Affirm: He could pay the $2,200 bill in 12 installments, at an interest rate he could live with. His home is in South Berwick, Maine, and “money doesn’t grow on trees up here,” he says. He clicked the buy button.
When the package from the jeweler showed up at his door, he says, it wasn’t the 1.27-carat cushion-cut diamond he’d chosen. “It looked like a Cracker Jack ring you wouldn’t pay 10¢ for,” he says. He sent the ring back and complained to Affirm. “Affirm only provides the funds for you for your purchase and nothing else,” a customer care representative wrote in a Jan. 25, 2019, email. “We do not deal with items, order, or merchant questions. You will have to contact the merchant directly.” The jeweler told him it never received the package. He emailed Affirm a picture of the FedEx receipt confirming the return was made. A different customer care representative wrote back, saying Affirm “ruled in favor” of the merchant “because the return was not made in accordance with the merchant’s policy.”
Leavitt refused to pay and Affirm charged off his loan in May 2019. He says it’s caused his credit score to drop 100 points, harming his ability to buy a house, and even blames the stress of the whole charade for helping to cause a stroke that’s left him unable to work. Matt Gross, a spokesman for Affirm, says the company temporarily pauses payments while it investigates disputes. “We want every consumer to have a positive experience with Affirm,” he says, “including when they bought the wrong item.”
Of course, merchants have been exploiting irrational consumer behavior forever. Lawless, the University of Illinois professor, and two colleagues conducted experiments a few years ago that showed the fiscal ramifications of humans being hardwired to avoid intense moments of displeasure. The “hedonics of debt—i.e., the prospective pain associated with debt” can lead people to make credit-related decisions that work against their economic interests. “As long as monthly payments are affordable,” Lawless and his co-authors, Dov Cohen and Faith Shin, wrote in a paper summarizing their findings, “prices can be increased because consumers are willing to make payments for a long, long time.” The paper refers to a seminal study on colonoscopies, which found that the duration of pain matters less than how painful something is at its peak and how painfully it ends. As Cohen puts it: “What holds for colonoscopies also holds for credit.”
Youth fades; we all have to get colonoscopies eventually. Even if a short pay-in-four plan or a longer installment loan doesn’t feel relevant today, BNPL providers see a future in which you’ll be a customer soon enough. Klarna and Affirm in particular have ambitions to enmesh themselves further into our fiscal lives. Klarna has a banking license in Sweden (though Siemiatkowski says in his Twitter bio that he’s “Trying my best to be the nightmare of the bank establishment worldwide!”) and offers savings accounts to people across Europe. A recent “innovation” is an option called Pay Now, which involves, as the name suggests, no credit at all. You just pay now. Affirm offers savings accounts backed by the Federal Deposit Insurance Corp. that accrue interest and from which users will soon be able to buy and sell crypto.
Born on the internet, the BNPL industry may see its future growth come from a far bigger market: people buying things in person in actual real life. Klarna now offers a physical plastic card, allowing you to pay-in-four not just via merchants with whom it has partnered but for whatever, wherever you’d like. Affirm has a card that can act like a normal debit card or enable you to split payments into four interest-free installments. Levchin has described himself as particularly enthralled that people are using it heavily at Walmart to buy groceries, suggesting it’s “top of wallet.”
Asked if he believes an impending recession will temporarily curb BNPL use or cause it to spike, Levchin demurs. “But I’m pretty confident that three years from now there’ll be a lot more buy now, pay later in the US than there is right now,” he says. His vision involves you using Affirm to buy everything, causing you to abandon credit cards altogether. “I’m extremely biased, and I’m sort of drinking my own Kool-Aid and trying to sell it at the same time here, but I do see the demand for this credit-card-alternative payment modality just rising among young people, among people who are fed up with the sort of ‘fine print is my business model’ approach that traditional banks have,” he says. “I don’t take it for granted that everybody has to believe in it. But as an entrepreneur it’s my job to believe that my future is the best one.”
It’s hard to argue against the world BNPL companies say they’re creating—one in which finance is friendlier, transactions simpler, terms more transparent. In the broader movement to add a certain aesthetic to lending, with vibe-y names, better-looking websites, smartphone apps, and a boatload of marketing, there’s something powerful in rebranding credit and debt as more palatable. It’s just that for the spenders, the risks and responsibility remain the same. —With Jason Grotto
©2022 Bloomberg L.P.