AFRM - When a Noble Business Model Means Losses

[Machetephil]

Chart:

AFRM

Content:

AFRM had their crazy earnings fiasco yesterday where some intern accidentally (?) posted a portion of the company’s earnings early on Twitter and AFRM then released its entire earnings report before market close. The stock tanked on the results.

The company reported a 2nd Quarter December 2021 loss of $0.57 per share on revenue of $361.0 million. The consensus estimate was a loss of $0.44 per share on revenue of $330.7 million. The Earnings Whisper number was for a loss of $0.46 per share. The earnings call was even worse. The company’s CEO and CFO kept saying the company was growing at a rapid pace and continuing to grow, but failed to adequately answer any questions whatsoever about how they will be able to trim operating losses.

Normally, when a company is scaling up and not yet profitable, you want to look at what the company’s expenses are, whether revenue growth can eventually outpace expenses, and the target revenue for the company to hit to become profitable. (SNAP is a good example of this, having finally achieved sufficient revenue to report its first positive EPS since going public). This is where the problem for AFRM lies and why I believe there is a strong medium to long-term bear case for AFRM.

Here is the report: https://investors.affirm.com/news-releases/news-release-details/affirm-reports-fiscal-year-2022-second-quarter-results

The company reported 27.55% revenue growth in the most recent quarter at $127 million, yet posted an operating loss for the quarter of $196 million. Some of the loss can be attributed to growing pains and certain expenses that are critical for rapid growth but will smooth out over time. For example, the sales and marketing expenses increased by 266.8% from this quarter in 2020. That could be attributed to rapid growth and would likely be expected to smooth out over time and decrease as a percentage of revenue. But that’s not what I am focused on. I am focused on the largest quarter-to-quarter increase in expenses for AFRM: provision for credit losses.

AFRM experienced a 320% increase (!!!) in provision for credit losses for this year compared to 2020, from $12.5 million to $56.2 million. More concerning, their provision for credit losses as a percentage of overall revenue increased from 12.5% to 41.4% (!!!). They had no explanation for this on the call or how they are going to fix it.

What does provision for credit losses mean? Well, when a company like AFRM extends credit to customers by letting them buy a product now and pay for it later, a certain portion of those customers will, for whatever reason, default on their payment obligations. Some of those customers who default will never pay the money to AFRM. Others will, but not until AFRM incurs collection costs.

Here’s the problem: AFRM consistently market themselves as a company that does not charge late fees ever, charges no hidden fees, and has many instances where they charge little to no interest. This is a noble endeavor, but how does a company do that and still make money, especially when some of its customers don’t pay back the loans?

The answer: they don’t. AFRM had to increase its provision for credit losses by 320% but has no solution for how to reduce its provision for credit losses. They continue to say they won’t charge late fees. They continue to offer zero-interest options. They continue to provide lower interest rates than credit card companies. Their stated solution is growth, growth, growth. But the credit loss reserves are outpacing growth and increasing exponentially both in raw numbers and as a percentage of revenue.

Unless, AFRM can get this growing problem under control, it continue to experience massively negative EPS in the near term. Seeing as AFRM does not even acknowledge this is a problem and, even if it did, its stated solution is growth, AFRM does not appear to be able to solve this problem with current management.

I believe that AFRM will continue to drop in value before leveling off to much lower valuation until and unless it can reduces its losses. I am going to look into puts on Monday after the Fed meeting.

[bajillion]

I’m watching this one as well. Your breakdown of their earnings validated my suspicion which is it’s easy to make money in lending when your borrowers pay your back. But most people using BNPL are credit risks and lots of them are high credit risks. There is a reason most micro-lending and payday loan businesses charge obscene interest rates. They have to offset a high default rate. If AFRM isn’t going to charge obscene rates they aren’t going to remain solvent very long.

[The_Ni]

AFRM experienced a 320% increase (!!!) in provision for credit losses for this year compared to 2020, from $12.5 million to $56.2 million. More concerning, their provision for credit losses as a percentage of overall revenue increased from 12.5% to 41.4% (!!!). They had no explanation for this on the call or how they are going to fix it.

That massive increase in credit provisioning in such a short period of time is extremely concerning. AFRM was supposed to be the best of breed for BnPl.

It is therefore likely that the following will suffer similar impairments, since they too increased their loan book significantly, and offer similar products:

• OPFI - targets similar customer segment. Earnings seem to be far away though (4/28), so we might not know how much more they are hurting, if at all, unless they adjust guidance earlier.
• KPLT - targets subprime customers. Interestingly, AFRM forwards to KPLT those customers who they deem too risky. So KPLT should be hit even harder. Earnings is 3/15, and has relatively liquid options.

Might nibble on KPLT Puts on Monday.

[Machetephil]

I tried KPLT puts on Thursday after early AFRM earnings release. I made out only very slightly green yesterday. KPLT seemed to be tracking SPY as opposed to AFRM. OPFI did seem to be quite a bit more down on Friday. I may look at some longer puts on these or some naked calls.

[rollover]

The past twenty-four hours have been a wild ride for Affirm (AFRM 46.55, -12.13, -20.7%), a provider of buy now, pay later services that issued a disappointing 2Q22 report yesterday afternoon. In anticipation of strong results, shares were in rally mode throughout the day. That rally kicked into high gear in the afternoon when some of the company’s results were mistakenly leaked on Twitter (TWTR), showing that GMV and revenue for the quarter surged by 218% and 77%, respectively. At that point, there was little doubt that the remainder of the report would be strong, especially since AFRM recently locked up an agreement with Amazon (AMZN).

That assumption proved to be wrong. Shortly before the close, investors had the rug pulled out from under them when AFRM’s full report was released. Almost immediately, the stock tanked, as a few glaring blemishes were revealed, including:

The company’s guidance was disappointing on multiple fronts. For Q3, its revenue outlook of $325-335 mln was merely in line with expectations, and its GMV forecast calls for a 19% sequential drop. Due to seasonality, a drop-off in GMV is expected, but this projected decline is substantial. For a reference point, GMV fell by 8% on a sequential basis in 1Q21.

AFRM only raised its FY22 revenue guidance by $60-65 mln to $1.29-1.31 bln, which is pretty weak considering that it exceeded Q2 estimates by $28 mln. In essence, the company raised its 2H22 revenue guidance by a paltry $35 mln. For many companies, any bump to revenue guidance would be considered a win, but expectations are much higher for AFRM given its landmark partnership with AMZN.

Compounding the problem, AFRM sees adjusted operating margin sinking to (21)-(19)% in Q3 and to (14)-(12)% for FY22. This forecast represents a major reversal from FY21’s adjusted operating margin of 1.6%. During the earnings conference call, CFO Michael Linford linked the margin guidance to increasing investments to grow the business.

On the topic of profitability, AFRM’s GAAP net loss of $(0.57) missed expectations as total operating expenses soared by 141% yr/yr. The company doubled its headcount and also ramped up marketing investments around the holiday season.

Rising interest rates are another concern, particularly regarding AFRM’s split-pay arrangement with Shopify (SHOP), which is expected to comprise 15-20% of GMV in FY22. Specifically, the problem resides in AFRM’s revenue less transaction costs metric.

Last night, Linford explained, “… taking on a Split Pay product with 5% to 5.5% merchant fees, you’re not going to be making 4 points of margin. And so you do expect a little bit of compression on a percentage of GMV basis on the Split Pay business…”

This headwind does seem manageable, though: Linford estimates that a 100-bps increase above the increase implied by the current yield curve would only result in a 10-20 bps impact to revenue less transaction cost as a percentage of GMV.

The bottom line is that an expensive stock like AFRM, with a mid-teens forward P/S prior to this crash, is very susceptible to a sell-off if the growth projection weakens – especially in this current environment.

[Machetephil]

What is strange about AFRM is that increased growth is actually worse for shareholders until/unless they get their operating expenses under control. This means any news about AFRM growth that results in a stock price increase is a good opportunity for taking a short position as next earnings will be just as bad if not worse.

[The_Ni]

To add to this, there are two lagging effects in play.

First, many don’t often default right away. They will make a few payments, then fall behind, and then eventually stop. So a good chunk of the loans that show up as current in month 1 can show up as delinquent in month 3.

Second, the level of provisioning depends on how late the loans are. I couldn’t find the aging schedule on the earnings report on their website and don’t remember the exact regulations, so lets use a stylized examples. Not 100% of the outstanding amount of a loan that is late is provisioned for - it depends on many days it is late. Loans that are less than 30 days late only may be provisioned at 25%, while those that are 180 days late may be provisioned at 100%. So that number that we are seeing - 41.4% - likely contains loans that will have to be provisioned for at higher levels as time passes.

This means that even if they were to hit the brakes now in terms of slowing loan origination, first late payers, and then provisioning will catch up to them and make their balance sheet get worse and worse, and potentially even blow up.

It would therefore actually be in their interest, perversely, to keep growing their loan book, as they get their house in order.

This is not new to payday lenders, of course. They way they mitigate this is by charging 200%+ APR, so that the good borrowers subsidize the delinquent ones.

[Eze123]

Lots of great insight here, I wish I would have been able to see such a convincing bear case a couple months ago

I cut the rest of my AFRM on Friday, short term I don’t want to try to guess a bottom. I did expect increased credit loss provisions/delinquency, but as others have pointed out, the increase this quarter was serious. On top of that, guidance was shit - though I will say their leadership has historically been cautious in this respect,

I do still think there is a serious long term case for AFRM, and given the cautious guidance I think they are set up to beat revenue and eps expectations in May. I suspect their operational spending, which was way up this quarter, will slow - less advertising spend post-holidays, less spend for building out/scaling to new partners/AMZN, and less R&D as they already released Debit+ and SuperApp.

But, short term they are boned in this economic climate. I will probably buy puts on KPLT, I strongly considered that in the past after hearing the news of their AFRM/AMZN partnership, and hearing they only took the subprime accounts. I hoped that companies like KPLT would absorb many of the problem customers from AFRM, and as we have seen there are plenty of them, so I don’t see how KPLT fares any better.

I think it is worth looking into SQ puts as well - having overpaid for AfterPay and likely weathering similar credit issues as AFRM. Though diluted against the rest of their non-BNPL revenue.

[rollover]

https://twitter.com/unusual_whales/status/1493072758627962885?s=20&t=X5ms5z1w1090LPlvTHDEjg

[The_Ni]

Went through their Q2 '22 earnings supplement, and it has some interesting information that seems to make the prospect of deteriorating portfolio quality less severe.

1. 30+ day delinquency is about 2% (slide 21). This is likely calculated as (amount overdue)/(total loan book balance), so the actual portfolio at risk is going to be a few times higher, but the graph below tells us that: a) there is a cyclicality to delinquency so it is likely to go down from here, and b) it’s actually lower than 2019 and 2020.

2. We see some of this cyclicality in the provisioning too, including a “de-provisioning” in Jun 2020 and Mar 2021, for example. (Slide 31)
   
   

Having said that, there are two things that jumped out that can give one reasons for pause.

1.The proportion of the most highly rated loans (ITACS Score of 96+) went down from 75% in Dec '20 to 57% in Dec '21 (slide 22). I don’t know what is in the ITACS Score, but they seem to tout it, so having to take this at face value.

2. Write-offs are 2% of the portfolio overall, but for the 2020 cohort, is actually 6% (slide 23). Bit of an uptick.

Net-net, I feel less alarmed than I was, but still feel caution is warranted. Another earnings cycle should make it clear if we are staying on the cyclical track, or if pressures on people’s wallets and a rapidly growing loan portfolio have combined to make things materially worse.

[Yongsooyuk6]

AFRM is currently up 65% since 3/14, or a week ago from today. Is there any reason for this and is there more room to run? Would going back to pre-earnings levels ($60-80) be possible?

[TheMadBeaker]

Not sure on the run up, there’s been a lot of mixed news. Some still very bullish, some still very bearish. I think for the most part it’s just sympathy movement with SPY.

[TheMadBeaker]

I’m joining you @Machetephil with May puts, I got a couple of the 35p for 3.55 on that pop to 43.25. If SPY tanks in the coming days it’s going to pull AFRM right down with it. After their last earnings fiasco I can’t imagine anything good coming from them for quite some time…

[nano]

Got a 3.55 fill as well.

[LexuSensei]

I got in a full for 3.55 as well

[LexuSensei]

Averaged down a little today at the peak, sitting at 3.35 avg price currently.

[nano]

Cut for 15% on the dip this morning, will look to re-enter if there’s a good opportunity. Thanks for the call-out @Machetephil

[TheMadBeaker]

Sold my longer-out position on Friday for good gains, thanks phil. Also when SPY decided to knife late morning from the Saudi Attack news, AFRM was the first stock I thought of and quickly grabbed a FD for some quick gains!

Going to wait for another good entry opportunity.

Lowest drop recently was: 25.26
Major support level: ~36

I think it has room and reason to fall again, especially if SPY decides to reverse course and head back down.

[Vincenzo]

What do you guys think of getting calls for next months earnings, it ran up during last quarters earnings. I’m thinking 8/19’s just to be safe

[Randall]

they actually missed last earnings hard (the one they published accidently mid trading day). i wouldn’t take risk on that firm when there are better plays to be made. (unless its for the gamble, with a small position).

also, rate hikes and inflation are super bad for companies that work based on sales commission, someone already said in VC that Mastercard and Visa plan to increase their fees and I don’t see AFRM as a company with significant pricing power that will be able to do such thing.