I’d like to post my thoughts on Affirm and how I came to the conclusion to increase my position after the latest earnings report. I originally bought in after the news of Afterpay’s acquisition by Square in August.
It’s definitely not the typical SaaS hypergrowth stock. So, everyone’s thoughts are greatly appreciated.
- Affirm is a tech company that facilitates Buy Now Pay Later (BNPL) transactions. Its underwriting technology is the central pillar underlying its business model and future success.
While the BNPL arena is very very competitive, Affirm has a strong track record of lower default rates thanks to being an early mover in AI/ML underwriting within BNPL.
The company was founded in 2012 and has had 8+ years in the space. I think this provides Affirm a defensible moat.
Per their S1, they stated their credit underwriting risk model takes “five top-of-mind data inputs and turns them into a total of over 200 data points in order to assess the credit risk of new consumers.”
Here you can see the DBRS Morningstar rating report on Affirm’s latest (Aug 2021) securitized trust: https://ds5q7tzjvgeff.cloudfront.net/document/382643.PDF?Exp…
If you view page 26, you can see the Cumulative Net Loss (CNL) rates were above 4% at the highest in earliest cohorts; the key here is the CNL has been marching steadily lower for each successive monthly vintage from 2016. Recent cohorts might even level out at 2% CNL or below.
The 2019 cohorts (which is precovid/prestimulus) is a key marker in my opinion, as they have CNLs performing quite well despite the pandemic shock, and better than earlier year cohorts (that were not impacted by the pandemic at all).
Of interesting note, this latest trust report shows 34.36% of loans carry zero interest at wtd avg 708 FICO, and at average loan size of $466.
Of the 65.64% of loans with interest the weighted average interest rate is 25.5% at wtd avg 668 FICO and at average loan size of $647.
(Per the last 10Q, 0% APR financing represented 43% of total GMV so the trust is weighted differently than the overall actual transactions)
During COVID last year, Affirm increased the allowance for credit losses as a percentage of loans held for investment from 8.9% in Feb 2020 to 14.8% in March 2020.
Since then, due to stronger than expected repayments, the allowance was released and as of June 30, 2021 the allowance was down to 5.8% (but upticked from 5.2% in the quarter prior).
Also lending credence to the idea that Affirm possesses a technological moat is the recent announced August 30 2021 Amazon partnership. Even though Amazon has had a way for customers to BNPL (a pay in 5 monthly installments option) it is very limited only to the best creditworthy customers.
Clearly, Amazon believes it is too costly and time consuming to build its own AI/ML models plus the years of training data, to match what Affirm already offers with its ability to select the right customers for BNPL and at the right interest rate pricing. Amazon must believe Affirm can expand its customer transaction volume with its payment options. The option will be available for above $50 order sizes.
And back in July 2020, Affirm announced an exclusive partnership to power Shopify’s Shop Pay installments in the US.
The company also made moves to acquire PayBright in Dec 2020, which is Canada’s leading BNPL provider for about 267M USD. In August 2021 it was also announced that Apple is expanding iPhone, iPad, and Mac monthly financing to Canada with Affirm’s partnership via PayBright.
Again, this supports the idea that Affirm’s tech differentiates itself from other competitors to win these partnerships.
2.) Affirm is disrupting an industry. This is a good sign, as disruptive companies are more likely to have high growth and a massive TAM.
The company’s mission statement is to “deliver honest financial products that improve lives”.
Founder Max Levchin wants to evolve payments from credit cards. He believes credit cards have ‘devolved’ over time with late fees, overdraft charges, 'gimmicks like deferred interest. His goal is to really disrupt the credit card industry.
In fact, Max Levchin has called its ongoing Affirm debit card roll out, announced back in February, as “the anti-credit card”.
Max wants consumers to shift to a transparent, clear form of payment: the ‘buy now, pay later’ model, which allows consumers to pay for purchases in fixed amounts over time without deferred interest, hidden fees, or penalties.
So depending on their creditworthiness, consumers might face 0% APR. Less creditworthy consumers would have fixed amounts of interest built into their purchase - interest which never compounds. There are zero hidden fees or costs. Even if a consumer misses a payment they will 'never pay more than agreed to at checkout".
Consumers can choose between biweekly, 3, 6, or 12 months term options that best fit their needs, on a per-transaction basis. [Other competitors don’t offer the same flexibility]
The value proposition for merchants is that Affirm can boost Average Order Volume (AOV, defined by GMV divided by number of transactions during the measured period) by 85%, increase conversion rates by 20% (reducing cart abandonment), and that 20% of Affirm consumers will make repeat purchases at the same merchant within 12 months which increases the LTV.
Since 2016, Affirm has processed over $17B in GMV and has 7.1M active consumers.
Now, Affirm is not a bank, although it can lend in certain states on its own.
Affirm partners with Cross River Bank, just like UPST (and Celtic Bank is their only other partner, with de minimis contributions).
CRB will originate loans for Affirm. CRB will offer the vast majority of these loans to Affirm for purchase. Affirm then buys back the loans. Affirm basically takes on the risk of payment default. Affirm does sell some of these loans to capital markets (as outlined by the securitized trust example above).
Regarding the TAM, online sales are expected to be about $1.32 trillion in the U.S. retail market by 2023 (Affirm does not currently operate outside of North America).
In 2020, BNPL accounted for a record $24 billion in U.S. spending. A Bank of America report from December 2020 predicted that the BNPL market could grow 10–15x by 2025 and process between $650 billion and $1 trillion in transactions annually. https://finledger.com/2020/12/30/special-report-the-buy-now-…
(Affirm also has a virtual card offering that can allow offline purchases as well, increasing its market opportunity)
3.) Affirm has stellar management.
The company is founder led, by CEO Max Levchin. He has of course a great track record, was cofounder of Paypal and its CTO.
Glassdoor ratings show 4.4 star of 195 reviews, and 90% approve of CEO.
Management has a strong customer focus, which always serves companies well in the long term (like Jeff Bezos/Amazon’s customer obsession).
There’s huge emphasis on customer service (see https://docs.affirm.com/affirm-developers/docs/customer-faci…).
They want full control of the customer repayment experience as evidenced by the fact they service every loan themselves (like with Upstart).
They recently acquired Returnly in April 2021 (for $300M) to make returns/exchange seamless and painless.
They also don’t just onboard any merchant, they will drop or deny merchants from joining Affirm’s Network if they don’t have the right fit (https://www.affirm.com/business/blog/why-we-can-t-work-with-… ).
Management’s goal is to create a strong defensible ecosystem and brand, that cuts credit cards completely out of the equation.
Their app personalizes brands, offers, deals to each customer to anchor them in repeat transactions. They now even offer a high yield savings account introduced last year - which attracted total deposits of approximately $300 million, with “no fanfare and no promotion, simply through organic engagement and a great product offering.”
4.) Affirm’s business model. How do they make money?
Merchant network revenue: Affirm charges merchants a percentage fee for every sale through Affirm’s platform (probably around 3% merchant discount rate). Fees though, are generally larger if the loan is 0% APR, which makes up for the lack of interest income collected by Affirm compared to less creditworthy borrowers.
Virtual card network revenue: Affirm takes a cut of an interchange fee if a consumer uses its single use virtual debit card to shop with a merchant not integrated with Affirm.
Interest income: Affirm makes money on interest bearing loans that they choose to retain.
Gain on sales of loans: Affirm might make money from selling loans. After Affirm buys back loans originated by Cross River Bank, it will also sell some of the loans to capital markets. This can be a gain or loss when done so.
Servicing income: Affirm also earns servicing fees for loans that were sold in the capital markets to third party partners
Affirm also will eventually get revenue from its ongoing Affirm Debit+ card rollout, but is currently immaterial. The purpose of the card is to convert any non-Affirm purchases already made by the consumer eligible into an Affirm transaction.
5.) The numbers
Q4 2021 ending June 30’s earnings report was this past Thursday Sept 9.
Gross Merchandise Volume (GMV) grew 106% YoY from 1.203B to 2.484B
Active consumers grew 97% YoY from 3.6M to 7.1M.
Active merchants grew 412% YoY from 5700 to 29000.
Revenue grew 71% YoY from 153M to 262M. This is 13.5% QoQ growth, in line with recent QoQ growth:
6.38% → 10.89%-> 13.5% → 17.3% ->13% → 13.5% for Q4 2021
But, we must factor in seasonality, so YoY rev growth might be better to look at:
89% → 109% → 98% → 57% → 67% → 71% for Q4 2021.
Looking at YoY there appears to be acceleration over the last 3 quarters, and it’s better to look at the recent quarters which are farther away from the distortions of the pandemic.
Revenue as percent of GMV fell from 12.7% a year ago, to 10.5%. This metric ranged from 9.7% to 12.7% over the past 8 quarters. This seems in line with expectations. I don’t expect this to range much above 10% particularly as SHOP merchants continue to integrate.
Overall, adjusted operating income as percent of revenue, is steadily improving over time as GMV scales:
(51.1%) → 30.5% → (4.6%) → 1.5% → 2.1% → 5.4% for Q4 2021
Average Order Volume (AOV) fell by 26.3% compared to a year ago, from $672 to $495.
I think some of this is due to increasing use of their Split Pay offering (a short-term payment plan for purchases under $250 with 0% APR, which increases the mix of shorter duration, low AOV products) plus I’m assuming the effects of Peloton price cuts.
2022 guidance also says that Split Pay offerings, which include the Company’s Shopify partnership, are expected to contribute 10-15% of the Company’s fiscal year 2022 GMV.
Total network revenue take rate fell from 7.3% a year ago, to 4.3%. Total network revenue take rate is calculated as merchant network revenue and virtual card network revenue, divided by GMV.
This metric over the last several quarters:
5.95-> 7.31-> 6.72-> 5.32 → 4.95-> 4.3 for Q4 2021
In the call, they say the elevated 7.31% take rate a year ago was due to “certain high AOV merchants elected to pay increased MDR (merchant discount rate, which is the merchant fee) in order to secure additional revenue in the early days of the pandemic.” However, the trend is definitely downward. I think we can expect further downward pressure over time as competition intensifies.
Sales and Marketing has increased substantially.
Sales/marketing (excluding stock based compensation) as percent of GMV has steadily risen over the last few quarters:
0.48% → 0.33% → 0.47% → 0.96% → 1.33% → 1.61% for Q4 2021
They had said in past 10Q “We increased our investment in sales and marketing channels that we believe will drive further brand awareness and preference among both consumers and merchants. Given the nature of our revenue, our investment in sales and marketing in a given period may not impact results until subsequent periods.”
But at the same time, the partnership with SHOP is merchant pays 5.9% of sale value split between SHOP and Affirm. It’s not clear how much goes to SHOP, as the amount is to be recorded under Sales and Marketing.
Probably very little of the current Sales/Marketing is from SHOP right now. So I’m actually skeptical Sales/marketing expense will significantly fall, as a share of GMV, over time.
The portion of interest income related to consumer interest payments grew 77% YoY.
The margin profile of the lending business seems to have improved over time. I derived this by calculating [interest income +loan sales+ servicing income] versus ‘total transaction costs’ [loan purchase commitment, credit loss provision, funding cost, processing/servicing costs]
Over the past few quarters it has progressed positively:
(127%) → 30% → (73%) → (16%) → 18% → 25% for Q4 2021
Cash: 1.46 billion
Balance sheet risk and capital intensive. Affirm takes on default risk for its loans.
Economic cycle risk. According to Credit Karma, a third of U.S. consumers who used BNPL have fallen behind on one or more payments, and 72% of those said their credit score declined. And this is in a benign credit environment today. Like with Upstart, a recession would definitely impact Affirm’s business, even if its underwriting is superior. Capital market fund availability would also drop precipitously. https://www.reuters.com/technology/buy-now-pay-later-surges-…
Regulatory risk. Affirm doesn’t have a CFPB No Action Letter for its machine learning /AI credit underwriting like Upstart. However, Affirm’s CEO has said he frequently engages with the CFPB and he has sat on its advisory board since 2015 (https://www.linkedin.com/pulse/why-im-joining-us-consumer-fi…)
Lumpy growth and seasonality. Similar to Upstart, it’s not a SaaS business. Per their S1, “we generally experience seasonality in our business in terms of changes in GMV in accordance with retail and e-commerce trends. We typically see increased revenues in the second fiscal quarter of each year as a result of the increased GMV occurring as a part of the holiday shopping season”.
Competition. Affirm used to dominate the BNPL market in the US in 2018, at 78% market share, but is now down to 16% in 2021. Competitors include Afterpay, Perpay, Klarna, Quadpay, Sezzle. And of course, credit card companies, Paypal, etc can also fight with their own BNPL solutions. Mitigating this competitive environment is a technological ‘moat’ that I believe Affirm has, and that the TAM is so massive for more than one player to thrive. https://www.emarketer.com/content/affirm-no-longer-dominates…
Bank partner concentration risk. This is actually more of an issue than with Upstart, funnily enough – almost all loans are originated via Cross River Bank. Celtic Bank is their only other partner, in their S1 they said Celtic was “de minimis” in originations. As I’ve posted before, CRB is currently under fire from Congress for poor fraud/verification of PPP loans last year but likely not an existential threat to the bank at this time.
Merchant concentration risk. This is rapidly decreasing. Peloton comprised 20% of total revenue in quarter ending March 2021, down from 28% in March 2020. Although, Peloton is cutting their prices and has had recalls recently which impacts this. In any case, total network revenue growth excluding Peloton was 106% YoY, faster than the overall growth rate (which includes Peloton) of 71%.
Fully diluted shares by treasury stock method at 9/8/21 closing price was 317M. Its market cap as of 9/10 close is nearly $40 billion. If based on their upper end 2022 FY guidance, that’s a NTM price to sales ratio of 33. Although, Afterpay was sold to Square at 23 times NTM revenue.
That being said, this could be considered a super rich valuation for a business that has no SaaS like margin profile.
However, 2022 revenue guidance is conservative, at 1.19B on the upper end. FY2021 they did 870M revenue. That’s 36.8% YoY rev growth. I think it’s definitely sandbagged like crazy. Their guidance is not including Amazon.
They stated on earnings call that they had on boarded 12,500 Shopify merchants in May, and today it is now at hundreds of thousands now on boarded. Apparently that is the stage before activating them.
Per earnings call:
“So just to give you a sense how the process actually works.
So you first have to educate the merchant base, then you have to onboard them, then you start activating them both for marketing and just exposure to consumers and then eventually start processing volume.
And so you can kind of think of it in these four stages. The number we highlighted in both Michael’s remark and mine is the tens of thousands of actives across our territories, but obviously Shopify question here.
If you look at the on boarded, which is the stage before that number is in the hundreds of thousands now.
And so we will start ramping the next stages of that process for the merchants, but it will take time.”
Is that all really included in 2022 guidance? Surely some of the GMV but probably not entirely. If, say, Affirm penetrates 5% of SHOP’s GMV in 2022, that could add 10 billion in GMV alone.
And according to fund managers surveyed by Credit Suisse, some believe Amazon’s partnership with Affirm will add about 20B GMV by 2024. (see: https://pbs.twimg.com/media/E-3-fh1XoAY_YIB?format=png&n…)
So, I think FY2022 might ballpark into 1.7B in revenue. That puts yesterday’s valuation at 23x NTM sales, for a business potentially growing revenue at ~100% YoY. And this all could mean 4B revenue by 2024. This combined with improving path to profitability as GMV scales has led me to double my position in Affirm (albeit it was a very small % size to begin with, due to Upstart crowding out my portfolio).