Introduction
Buy Now Pay Later, better known as BNPL, has taken the world by storm. While BNPL companies have been growing and raising meaningful capital for over a decade, it is only in the last couple years that the term has really entered mainstream parlance. That should come as no surprise, considering that the BNPL market is a nascent market that is growing at a rapid pace, with transaction value doubling from $159B in 2021 to $316B in 2023, and an expected increase of more than 40% to $452B in 2027 according to Worldpay’s Global Payments Report.
The growth of BNPL makes sense when considering all that it has to offer to customers. Instead of having to pay fully upfront to purchase goods, customers can instead essentially take a loan from these BNPL providers, and pay back the installments with no fees, which can stretch out for up to three months. For example, you could either purchase a good for $300 upfront, or three $100 payments spread out over 3 months.
Sounds familiar? That’s because it's actually not all that different from credit cards. The main difference is that BNPL tends to be easier for subprime borrowers to get access to, due to a relative lack of hard credit checks compared to credit cards, as well as it being cheaper to spread payments out over time with no interest accrued on that unless you pay late. It is also an extremely important tool, as the middle class has become more strained with increasing inflation, and credit card defaults have soared above pre-COVID levels.
Even at first glance, there are several red flags regarding the BNPL industry. Offering unsecured loans to subprime borrowers is very risky. After all, the 2008 financial crisis was started off by defaults among subprime borrowers, but at least those mortgages were secured via property; in this situation, the BNPL loans are not even secured. On top of that, due to the nascency of the market, there isn’t stringent regulatory oversight as of 2024, nor a strong regimen of consumer protections for BNPL debt.
Keeping these aforementioned issues in mind, it becomes important to explore how this market needs to evolve in order to ensure that this doesn’t become yet another debt bubble to burst as the total volume of BNPL debt continues to increase. The BNPL market is still nascent and not too complex yet; however, it is crucial to understand potential red flags as the market deepens. This article aims to explore how the lessons of the mortgage crisis can help inform how to proactively stabilize the rapidly growing BNPL market.
The Current Situation of BNPL
There are a number of prominent BNPL providers, with some of the biggest being Klarna, Affirm, and Afterpay, which is owned by the large fintech Stripe. There are also BNPL arms of existing payments players, such as Visa, Mastercard, PayPal, and Chase. Apple is a company which recently entered into the market with its Apple Pay Later solution releasing in April 2023.
As previously mentioned, one of the main draws of BNPL is that there isn’t any interest paid by customers. Instead, merchants pay the fees, which are around 2% to 7% of order value. If a borrower is late on their payment, there are interest charges or late fees that are usually instituted.
In this regard, there is a lot of similarity between credit cards and BNPL. In most cases, interest or late fees are only an issue if one does not pay on time. However, a major difference is that when making a large purchase that one cannot pay off by the end of the month, credit cards will accrue more interest whereas BNPL allows consumers to pay it off even over a 12 month period. This is particularly important - it’s no secret that consumers are being squeezed through inflation. To make it even easier for consumers, BNPL options are easier to qualify for unlike credit cards, which may cause the customer base for BNPL to become more subprime borrowers.
These BNPL providers are issuing huge volumes of loans - for example, Affirm issued $7.5B in loans in Q2 2024 with 14.7 million customers, which comes out at around $1.9B lent every single month. This is a 32% growth from the $5.7B loan volume earned in the same quarter a year ago. If the growth forecast for BNPL market size ends up being accurate and Affirm retains its market position, they would be lending an astonishing $2.7B every month in 2027, or around $32B every year. With such large loan volumes already, it becomes crucial to take a deeper dive into funding sources. A large portion of BNPL providers fund themselves with debt, Klarna makes use of customer deposits. Affirm also has a significant portion (~30%) of its funding coming from securitizations.
The securitization of BNPL debt is a significant cause for concern, and companies such as Affirm are doing it continuously. It raised $500M in 2021 from an asset-backed securitization, and raised an additional $400M securitization in 2023. After this issuance, Affirm noted that there was strong investor demand for their products and that they have seen a “healthier tone” in the securitization market. As total loan volume increases, there is likely to be an increase in the complexity of financial processes within the industry, such as securitizations and derivatives.
To understand why this will be a concern, we need to understand the leadup and ultimate downfall caused by the 2008 financial crisis.
The Origins of the Mortgage Crisis
After the economy started to recover from the Dot Com Bubble post 2001, the economy of the US and several other countries saw sustained and bombastic growth, particularly within real estate. Incomes were going up across the board, and these good times resulted in home prices going up for Americans. Total mortgage debt in fact increased from $7.5 trillion in 2001 to $14.6 trillion in 2007.
A large reason for such a rapid increase in total mortgage debt was due to the rise in subprime mortgages. Subprime borrowers are those who have lower credit quality and are riskier to lend to than prime borrowers. Lenders were then compensated for taking on such risk via higher interest rates. While there was always risk of these borrowers not being able to make their payments, as long as the value of the underlying real estate held up, these mortgages were sufficiently collateralized for the lender to recover their principal amount through foreclosure.
Most notably, oftentimes those who originated these mortgages didn’t actually have skin in the game. By securitizing mortgages, a large number of these loans would be pooled together and then sold to other investors. This divorced these originators from the risk of the products themselves, since they would just sell them to someone else. As such, there was less incentive for originators to be very disciplined with underwriting, and instead a focus on loan volume was more profitable as they were just taking fees while risk was taken on by actual investors. However, it is important to note that risk retention rules now typically require the originator to hold a 5% economic interest (as defined by NPV of cash flows to the residual tranche holder divided by total notes issued) in the securitization.
The Increasing Complexity
The excesses of the mortgage market were not just confined to originating too many risky mortgages. As bundled up mortgages, called mortgage backed securities, were sold to investors, these same MBS were bundled up into specialized investment vehicles called collateralized mortgage obligations, or collateralized debt obligations (CDO). The CDO essentially held these mortgages, receiving principal and interest payments, and then paid these funds to holders of CDO debt. CDO debt was split up into different tranches - the senior tranches were more secure and low-yield than the underlying mortgages, while the junior tranches could get higher yield at the cost of taking losses before the senior tranche. Securitization was already widespread for more prime mortgages - the securitization rate of these mortgages saw a slight increase from just under 75% in 2001 to a bit over 80% in 2006. However, subprime mortgage securitizations increased from around 45% to 80% over the same time period, according to data from Brookings Institute.
Another major proliferation of this pre-2008 era was the credit default swaps. Contrary to what some people who watched the famous movie The Big Short may think, the CDS was not pioneered by Micheal Burry just before the financial crisis, but had existed for quite some time already, with the first major use being by the bank JP Morgan in the 1990s. In essence, a credit default swap is like insurance on a debt instrument - premiums are paid to the seller of the CDS by the buyer, and the seller makes up for any loss in case of a default on the underlying credit. However, the crucial difference between CDS and actual insurance is that the buyer of the CDS doesn’t actually need to own the underlying credit they are buying protection for - they can thus speculate on debt that they don’t even have using these naked CDS.
This became crucial to the subprime mortgage market by completely amplifying its size in an artificial manner - it was now possible to create synthetic CDOs. In a regular CDO, buyers put up money to buy certain tranches, and receive interest payments and principal payments over the life of their bond. A synthetic CDO pretty much offers the same function without the buyer owning the actual mortgages. In this case, the buyer actually sells CDS on existing CDOs. Instead of receiving interest payments, they receive premiums from whoever took the other side of the trade, and if the underlying CDO tranche defaults, they need to cough up the money to pay for the loss.
Theoretically, of course, this is not too functionally different from a regular CDO investment. However, what this means is that on $100 of mortgage debt, there could be even $500 or more betting on whether the debt is paid back or not. This poses very obvious risks - ISDA data shows that in 2007, credit default swaps were insuring an astound $62 trillion of underlying credit. There is no doubt that even if only a modest portion of this number was insuring mortgage backed securities, or MBS-backed CDOs, this is still multiple times higher than the actual total amount of mortgage debt outstanding. Suddenly, the mortgage market went from $14.6 trillion to at least double that value by even the most conservative estimate, all in a synthetic manner.
The mortgage market went belly-up from 2007-2008, resulting in widespread defaults across the US and subsequently the entire global economic system. The scale of the default crisis resulted in MBS holders incurring huge losses, with even holders of senior tranches in CDOs making losses. Perhaps the market could have withstood this - the major sellers of CDS insurance such as AIG would have taken on very heavy losses, but the world might have gone on after an incredibly bad financial year. However, the synthetic debt referencing mortgages meant that this was a cataclysmic event for the entire financial services industry. AIG, whose executives boasted that there’s no way they could even lose a dollar in CDS contracts, alongside other sellers of CDS collapsed trying to pay off all claims, being forced to sell off all the financial assets they held in order to repay claimants. Investment banks, who weren’t subject to stringent capital requirements and thus incurred maximum leverage while trading these mortgage backed securities, saw both demand for the products they were holding on their books and the ability to roll over and refinance the leverage they had completely dissipate, leading to the collapse of major institutions like Bear Stearns and Lehman Brothers. The sheer interconnectedness of all these parties led to a broader contagion effect, with this collapse rippling throughout the entire financial services sector and consequently the entire economy, taking years for the economy to fully recover.
The Connection to BNPL
The previous two sections on the debt crisis are not some sort of revelation - all that information is quite well known. However, it is important to tie those learnings to the BNPL market. When comparing the BNPL to subprime mortgage market, it is as if we are in a pre-2004 stage right now. BNPL volumes aren’t anywhere big enough to have very widespread effects on the economy, but the rapid growth means that total BNPL debt will increase by a lot over the next few years. Economic statistics may be all over the place, but middle class people on the ground are all feeling the chokehold of prices going up year over year, while wages are nowhere near able to cover this inflation, completely eroding purchasing power for the average person.
There were several key factors that made the subprime mortgage crisis so much worse than it could have otherwise been:
1. Use of derivatives to synthetically enlarge the market
2. Misaligned incentives between originators and investors
3. Perilous funding structures
These all run a very real risk of potentially being repeated in a few years.
For example, there is nothing inherently wrong with writing CDS on BNPL debt. It is a useful way for investors in the asset class to adjust their exposure to default risk, and this flexibility can allow more investors to be comfortable with the risks inherent to the space. However, if naked CDS proliferate within this space, that is a huge cause for concern. While both BNPL debt and subprime mortgages targeted risky consumers, BNPL doesn’t have any collateral per se, and is therefore even riskier. Additionally, BNPL borrowers are mostly less-wealthy consumers - if a large scale crisis does occur, the government is unlikely to allow BNPL creditors to squeeze every penny in recovery off these borrowers through the different types of recourse available in personal insolvency cases. Artificially amplifying the total referenced debt would be a ticking time bomb, especially coupled with rapidly ballooning loan volumes.
Additionally, the use of further non-equity funding sources, such as securitization, layers on an additional level of risk. Unlike the use of customer deposits, where there is a certain degree of further regulation, in a securitization, there is merely a requirement of 5% risk retention - assuming a wide enough excess spread in the securitization structure, it’s possible to hit this number with relatively little overcollateralization, which reduces the level of interest the originator would have in ensuring tighter underwriting standards. Right now, securitization in the BNPL space is quite low. However, there is nothing stopping BNPL vendors from tapping into securitizations as a form of selling their loans, which could loosen the underwriting that the vendors do, causing an enormous pile of very bad loans to enter the markets, highly sensitive to macroeconomic factors.
There is a lot of incentive to do this too. In the past few years, public valuations of BNPL companies has decreased substantially, decreasing the likelihood for BNPL companies to raise funding from equity issuances. Affirm, Block and PayPal, all very large BNPL providers, have seen their stocks drop by more than 50% in the past three years each. This decrease in investor sentiment, paired with the amount of competition in the space, may push these players to find new ways to compete for fundraising dollars. Furthermore, Klarna, which was once Europe’s most valuable startup, has rumored to be considering an IPO in late 2024 at a $20B valuation according to Bloomberg. This is something that investors are already hesitant of, as there are few investors in today's market looking to invest in a company that is barely profitable, with Klarna reporting its first quarterly operating profit in November 2023. If their IPO does not go according to plan, they too will look for new ways to outcompete others and raise additional funds to keep themselves afloat.
Finally, there was excessive leverage used by debt investors and traders, as well as an untenable volume of unfunded CDS written by counterparties. While stronger regulatory capital restraints have worked to keep banks in check, this same regulation has not been put in place for BNPL providers. As of 2024, the United States has yet to impose widespread regulations on BNPL providers, simply issuing a report in 2023 stating that they plan on increasing regulation in the future. The same vagueness applies to Canada, where the FCAC is currently monitoring the evolution of the BNPL market. The United Kingdom, however, has taken a stronger stance on BNPL. In February 2023, they drafted legislation which would give the Financial Conduct Authority (FCA) power to regulate firms offering BNPL services, and protect consumers from easy-access lending. This shows that there is precedence set with regulating BNPL providers, and sets the stage for regulation by giants such as the United States.
Conclusion
Buy Now Pay Later financing solutions offer a convenient credit option for consumers which benefit businesses and ultimately benefit investors through their securitization. While there are obvious connections between BNPL and the subprime mortgage backed securities market, the BNPL industry is a mere $316B versus the $685B mortgage backed securities market that crashed the global financial system, whereas the vast majority of this $316B is not securitized. However, as it is forecasted to grow 40% by 2027 and with increasing reasons to pursue securitization as a method of fundraising, BNPL companies possess increasing responsibility in a financial environment that appears unstable and uncertain.
If we look at the positives, BNPL can benefit consumers and businesses if properly regulated. A Bain survey stated that 40% of consumers mentioned that the ability to spread out the cost of a purchase was a top three reason to use BNPL services. Among financially struggling individuals, 81% of them stated that BNPL has made it easier for them to buy things that they need and only 44% of them stated that BNPL has made them more likely to buy things that they don't need. This signals that a flexible payment option with relatively low downside for consumers is beneficial to many people that struggle with their needs, not their wants.
A similar survey by Bain found that the average revenue per order using BNPL solutions is $114, which is much higher than the average revenue per order of $98.50 from credit card users. This, alongside new customer acquisition and higher basket conversion rates, indicates strong reasons for businesses to offer BNPL solutions, which is leading the industry to grow further and increases the risk of BNPL providers choosing to securitize these loans.
There is clearly a place in the market for customers and businesses to benefit from BNPL services, but the rising number of users, rapidly growing market share and option for subprime securitization leads to concerns eminent of the 2008 financial crisis. BNPL companies must be forced to disclose the credit situation of their customers and limit loans based on prior obligations, or else a rapid decline in the global economy could follow the monumental rise in BNPL.
