The Investment Case For Buying Affirm Shares Now (NASDAQ:AFRM)

Setting the table for a nuanced analysis of Affirm’s quarter and outlook

The last two days of the prior week were one of those periods that tried the souls of investors in high-growth IT shares. The inflation print was 0.1% above prior consensus expectations, and initially stocks fell, before sentiment reversed, and stocks started rising. Then one of the regional Fed Presidents spoke and sounded a hawkish tone with regards to inflation and the course of rates. The market proceeded to swoon.

Then came the President’s National Security Advisor to proclaim the possibility of a Russian attack on Ukraine. The uncertainty of what the Russians might do, and how we might respond and how that might upset geopolitical balances further unhinged fragile market sentiment. There are more than a few unknowns and unknowables in the firmament.

This is an article about Affirm (AFRM) and not interest rates or inflation or the denouement of the crisis in Ukraine, although one of the concerns investors have with regards to Affirm relates to interest rates and how they might impact the company’s business model. The fact is that higher rates are embedded in Affirm’s guidance, and that even higher rates will not have a material impact on the results of the company.

As most readers will know, Affirm reported the results of its quarter on February 10th. Some of the results were reported early, and the shares initially strengthened. When the entire release was published, the shares imploded. The shares lost 21% in the wake of the release and have continued to fall. Overall, the shares have lost about 40% from the levels that had been reached on the day prior to the release. It might be noted, that in the days prior to the release, the shares had appreciated by about 15% as speculation abounded with regards to the strength of the quarter that has now been reported. The shares have now made a low for the year, and a low since the company’s IPO that took place in January of 2021.

What should investors do at this point? Were the results or the guidance at levels that could reasonably have led to a 40% valuation implosion. Just what are the prospects for Affirm’s business? How much credence should be afforded to Affirm’s guidance? Are Affirm’s opportunities being appropriately assessed when looked at carefully and in context?

I own Affirm shares and have done so for quite some time, although I did sell part of the position when the shares spiked in the wake of its announcement of a partnership with Amazon (AMZN). With the shares down more than 70% since that time, and with the prospects of the company, carefully evaluated, stronger than has ever been the case, I plan to add to the position. I think the current valuation is exceptionally unrealistic and ignores both the performance of the company in this latest quarter and its record in terms of providing guidance. My time horizon is always at least a year, and my analysis is based on that timeframe.

Some analysts talked about Affirm shares “being in the penalty box.” I won’t suggest that won’t be the case… until the next event comes along that might refocus investor attention not on some of the issues I will discuss in detail below, but on the company’s opportunities, and its record of accomplishment.

Much of the analysis I have seen from brokerage commentary is neither nuanced nor, to put it bluntly, particularly analytical. But overall, analysis runs the gamut, from one comment from the Morgan Stanley analyst, “that the quarter was all that could be wished for,” to a downgrade from the Jefferies analyst. It has been the nature of this particular phase of the market to shoot first and ask questions later. The fact is the quarter that Affirm reported was staggeringly positive. The guidance the company has provided is, as is the wont of this management, exceptionally conservative. It should be considered, I believe, both in the context of the history of this management in providing expectations, and also in the context of just how strong fiscal Q2 actually was.

Just what did Affirm report for its fiscal Q2?

Affirm’s Q2 really was stunning in terms of revenue growth, substantially exceeding prior expectations. And for what it is worth, when carefully considered, the company’s cost metrics, and its gross margin equivalent were quite strong. This is particularly so, as this quarter lapped a quarter in which revenues from Peloton (NASDAQ:PTON) were particularly strong and had an outsized impact, both on revenues and margins.

I am going to confess that the foregoing part of the article was a bit tedious to write, and I am sure it will be tedious to read, but to understand what is going on at Affirm, it is necessary to understand the mix shift, and how that is impacting the recognition of revenues and costs. Yes, there is what appears to be a disconnect between the growth of GMV and reported revenues. But no, this does not mean that the business model is deteriorating, or that the growth potential for the company is less than thought or that the Amazon deal creates GMV and not revenue or margins for Affirm.

I also have to note that the CFO, in his conference call narrative, tried to cover this. When I wrote above that analysts weren’t analyzing, this is what I felt was missing in much of what I read. If analysts are going to provide advice and recommendations for clients, then understanding changes in product mix and their financial implications would seem to be a priority. But the fact is that much of the disconnect was simply ignored or said to represent some existential problem for the company. It is anything but, but disentangling the strands of revenue can be tiresome.

The headline revenue growth number for Affirm was 77%. Perhaps a better metric to look at would be revenues less transaction costs which grew by 93%, both of these metrics considered on a year-to-year basis. Some other revenue related metrics of interest: Customer growth on a sequential basis was 29%. The average number of transactions per average customer rose 15% year on year. The number of merchants on the platform rose by about 65% sequentially, reflecting the continuing progress the company has made in recruiting Shopify (NYSE:SHOP) sellers to offer Shop Pay.

GMV rose by 115% year on year. It reached $4.5 billion, up from $2.1 billion in the year earlier period. GMV excluding Amazon actually doubled, compared to 84% growth last quarter. Sequentially, excluding the contribution of Amazon, GMV rose from $2.7 billion to $4.2 billion, or more than 55%. That kind of sequential increase compares to 40.5% sequential growth between Q1 and Q2 the prior year. Consumers and merchants are voting with their wallets and adopting the Buy Now/Pay Later paradigm as an alternative to high interest credit card debt.

Why the substantial disconnect between revenue growth, revenues less transaction costs and the growth in GMV? As compared to some companies, Affirm has what might be considered as a relatively complex revenue model. Its revenues consist of interest income, which itself consists of 4 categories. The other revenue buckets the company reports include merchant network revenue, virtual card network revenue, the gain on sale of loans and servicing income. Revenue as a percent of GMV fell from 9.8% last year to 8.1% this year, and it fell from 9.9% the prior sequential quarter. That sounds bad, and some commentators have pontificated on the subject. But the reality is something very different.

A year ago, much of Affirm’s business was composed of the fees it received from its merchant partners for the “0%” rate offers. Those 0% rate offers were made not just by Peloton, but a whole host of merchants who used the 0% rate as a promotional device. The fact is that this component of the business is falling as a percentage of the total. Revenues from the merchant network services, which is where this revenue appears, “only” grew by 27% year on year, and by 38% sequentially.

A far greater proportion of Affirm’s business is now composed on interest bearing loans. Some of these loans are packaged and sold. That is shown as the gain on sales of loans and that revenue classification actually rose by more than 4x year on year, and by 84% sequentially. In the year-earlier 2nd quarter, there was no sequential growth in that revenue category. By now, however, interest income is Affirm’s largest revenue category, and accounts for 38% of total revenues. Interest income is primarily composed of the interest that Affirm earns on the loans held in its portfolio. Those loans are now a bit more than $2.4 billion, up by 29% year on year, by 20%+ since the start of the fiscal year, and by a bit less than 10% sequentially. That in turn led to growth of 87% in interest income year on year, and by 18% sequentially.

One of the factors leading to the disconnect between growth in GMV and growth in revenues has been the rise of Split Pay, which is the service offering currently used by Shopify’s merchants and a number of other partners. The Split Pay offering grew 4 times year over year and it is now more than 30% of GMV, about double the percentage in the year-earlier period. While the take rate on the Split Pay offering has remained consistent, because of the nature of the revenue recognition model, as the Split Pay offering has grown both in absolute terms, but also as a percentage of Affirm’s total, it has depressed the optics for current period take rates. Over time, of course, the take rate for Split Pay transactions, as can be seen on pg. 13 of the most current investor deck, has held fairly constant and ticked up last quarter. I really think before shooting the shares, it would be reasonable to look at the absolute levels of the company’s take rate over the last several quarters.

Affirm is partnering with larger merchants now; of course, Amazon is the largest merchant of all followed by Walmart (NYSE:WMT) and Target (NYSE:TGT). Needless to say, larger users are going to receive better terms than smaller merchants, and that too is showing up in some of the numbers that the company reported and is projecting. That said, however, the volume brought by these large users to Affirm’s platform is and will continue to provide positive leverage to the ratio of revenues less transaction costs.

In the latest quarter, Affirm sold 59% of the loans it generated in. But on the other 41% of loans held in the company’s portfolio, the company recognizes the revenues ratably, which means that compared to the mix of business in the past, revenue recognition is significantly deferred. On the cost side, the company makes a provision for credit losses up front. Thus, the reported margins on the loans that the company originates and holds in its portfolio are very back-end loaded. When loan volume is rising rapidly, as is currently the case, the reported gross profit from a substantial portion of business is deferred and will be recognized in subsequent periods.

Going into the release of quarterly numbers, some analysts were concerned about delinquency rates and loan losses. The raw numbers might look concerning, as the provision for credit losses in the quarter rose from $13 million a year ago to $53 million, this past quarter. That looks bad, but again, like many things, it is important to understand the context. In the year-earlier period, the company released $39 million of excess loan allowances that had been taken in the early stages of the pandemic. So, there was no substantive change in loss provisions year on year. In other words, what looks like deterioration was actually an improvement.

I don’t pretend to be a credit analyst. But a significant component of Affirm’s business is underwriting loans. I believe, and have done so essentially from the start, that Affirm has a substantial advantage over the other BN/PL competitors in regards to the technology it has to underwrite loans. There are, apparently, some BN/PL competitors who don’t underwrite the loans they make.

Some of the analysts who cover Affirm like to look at the results of the different tranches of credit that the company has sold. One of the things I have learned over the past decades is to avoid postulating conclusions with only partial visibility. There are some loan tranches whose performance is worse or better at particular times. But the results suggest that the secular trend of Affirm’s loans is not negative, and indeed, the company has a higher proportion of highly rated loans now, than it has had in the past. This data can be seen on pgs. 21 and 22 of the current investor deck.

Affirm is run by some of the sharper data scientists on the planet and that includes the CEO. It underwrites the loans it provides to consumers using an elaborate process with numerous interrelated regressions/correlations. During the conference call, the CEO chose to discuss that element of the company’s business in abundant detail. But the net is that Affirm manages to an optimized default rate, and as the overall economy and in particular employment has improved, it has relaxed some of its credit parameters. That is intentional! Again, the infographic presentation presents in graphic terms a picture of delinquencies that is pretty much unchanged across the different kinds of credit this company extends.

The fraught question of guidance

This is the 4th time Affirm has reported its results. Each time the company has reported, investors have been spooked by its guidance. When the smoke cleared after the following quarter, the company had exceeded guidance by some significant amount, and forecasted another quarter in which growth was supposed to slow to unacceptable levels.

I cannot speak to the motivations of the management of Affirm in making the projections as they have. I can suggest that many analysts, who make a living tracking such things, ought to be aware of the company’s track record and react accordingly when the company issues guidance that is not consistent with the qualitative component of its presentation.

As mentioned, the Affirm revenue machine was on full and gaudy display last quarter and some of the statistics with regards to year-over-year growth in particular segments really were stunning. The percentage growth specifics can be found in the conference call transcript, but suffice to say there was growth of between 200%-300% in a number of merchandise/service segments.

Overall, reported revenue, exclusive of the contribution from Amazon, apparently rose by almost 60%. This was substantially greater than the prior consensus. Overall, revenues exceeded the prior First Call consensus by about 10%.

The company’s guidance for the current quarter is for revenues of $335 million. Just how serious, and how much reliance ought to be placed on that guidance? In a couple of words, not much. There are two reasons for that assertion which perhaps may strike readers as unwonted arrogance.

The first of these is the ramp to be expected for Amazon’s partnership with Affirm. As best as I can determine, Amazon probably accounted for about $15 million in Affirm revenue last quarter. (Affirm has indicated that it gets a merchant fee of about 5% on Split Pay transactions and the Amazon contribution to GMV was apparently $300 million,) But that revenue was essentially all achieved in just 40 days, as the roll-out of the Affirm service on Amazon did not take place until just before Black Friday. Obviously, the 40 days in which Affirm generated revenue from its Amazon partnership are the 40 strongest days for retail in the calendar encompassing Black Friday/Cyber Monday and the heart of the Christmas selling season.

On the other hand, Amazon shoppers were just introduced to the Affirm payment alternative last quarter. As an Amazon shopper, I can state without equivocation, the Affirm payment button is not particularly easy to find if you aren’t looking for it. It is not on the basic check-out page. The adoption of Affirm payments by Amazon shoppers will take time and will be a ramp. Probably a fairly steep ramp, but one that just started in the last 40 days of Affirm’s fiscal Q2.

The largest component of revenues from the Amazon relationship is going to be interest income. While some interest income will be recognized through the sale of loan tranches, much of it will be recognized ratably as I described earlier. All of this suggests to me that sequentially, revenues from Amazon are likely to at least double, and more likely triple sequentially, and I doubt seriously if that is really a component of the current “guidance” for quarterly revenue.

The other issue is that of seasonality. Affirm is a company that finances consumer purchases. Consumer purchases typically peak in Q4, and then recede in Q1. Last year, the company reported revenues in its fiscal Q2 of $204 million. It then guided for revenues of $195 million for Q3. That perhaps seemed reasonable at that time, but in fact, Affirm’s Q3 revenues were $231 million, and that was despite a $3.5 million revenue reversal because of the recall of Peloton treadmills, as well as the headwind posed by a decreasing growth from Peloton, at that time the company’s largest single customer. Most companies try to set forecasts that they expect to exceed. Exceeding a forecast by 20%, however, is more than a bit unusual and should be considered when trying to reach a reasonable conclusion with regards to what revenue growth ought to be anticipated for Affirm.

Over the Black Friday/Cyber Monday period , Affirm was thought to have financed about 1.6% of total online transaction volume in the US. It seems quite reasonable to believe that this proportion will continue to rise both in the current quarter and over the next few years because of the partnerships the company has negotiated and because it offers consumers far more choices than other BN/PL service offerings.

Putting together historical data and the likely growth of Amazon as a revenue source would produce a revenue expectation that would show positive and not negative sequential revenues. At the least, I would be inclined to suggest that revenues for Q3 will be 10% higher than the current guidance that has been anticipated by the revised consensus.

The other major issue is that of margins. Overall, the company’s margin forecast is unchanged. Despite some headlines to the contrary, Affirm forecast that its adjusted operating losses for the fiscal year will be 12%-14%, identical to the operating margin loss the company projected the prior quarter.

Because Affirm is a financial company, some of the metrics it reports are not consistent with those most frequently seen in earnings reported by enterprise software companies. Thus, it can be difficult to contrast and compare Affirm’s business model with that of other companies – there really are no reasonable comparables.

But there are two major cost buckets to consider. It may come as a surprise to some readers considering the performance of the shares and the headlines about margin deterioration, but the company’s equivalent to gross margins is improving. Transaction costs, which are equivalent to the cost of sales for other companies, despite the impacts of mix shift, has actually been improving and has been doing so at an accelerating rate. Last quarter, transaction costs were 49% of revenues, compared to 56% of revenues in the year-earlier period and 60% in Q1 of the current fiscal year. Looked at another way, in the quarter that Affirm just reported, sequential revenues rose by 34% and transaction costs rose by 10%. That kind of improvement occurred despite the mix change which typically defers revenue and increases costs in the short term.

In fact, as mentioned earlier, Affirm’s results in Q2 FY ’21 benefitted significantly because it recaptured $39 million of provisions it had reported at the start of the pandemic when the company was concerned that some of the credit it advanced would be adversely affected. Adjusted for that one-time benefit, the improvement in the equivalent of gross margins was actually from 73% in Q2 of FY ’21 to 49% in the just reported quarter. It seems fairly evident that despite the implosion of the share price, the reality is that the equivalent of gross margins for Affirm is showing substantially favorable trends primarily due to leverage at scale.

Affirm has 3 major categories of operating expenses. These are Technology and Data Analytics, approximately the equivalent of research and development, Sales and Marketing, and General and Administrative costs. These 3 items aggregated $193 million, or 53% of revenue last quarter compared to $92 million, or 45% of revenue in Q2-FY 2021. That kind of regression is obviously of concern, although to an extent, some of these expenses reflect the investments that have been necessary to launch the numerous partnerships including those with Amazon, Target, Walmart, Shopify and many, many others.

The picture looks quite a bit different on a sequential basis. Operating expenses in fiscal Q1 were $158 million, or 59% of revenues. The sequential growth of non-GAAP OpEx was 22% compared to the growth of 34% in revenues. Given the rapidity with which the company has been introducing new services, and its continued focus on improving its underwriting models, investors should want to see the company spend on Technology. And given the need to acquaint consumers with its services, spending growth on sales and marketing is an imperative. The return on these investments is likely very high, and the concern of some analysts about the company’s ramp of OpEx, at least to this writer, seems to reflect a very short-sighted perspective on the opportunities the company has.

Operating margins are obviously very dependent on the growth of revenues. If Affirm achieves the level of revenue growth that I think is far more reasonable to expect, rather than the guidance it provided, then the progression of margins is going to be far more positive than the 12%-14% operating margin loss the company has forecast. There really is not enough information to try to make any kind of well supported projection with regards to the exact trajectory of operating margins.

The company has forecast Q3 revenues of $335 million, and $190 million of transaction costs. Taken on its face, that suggests that revenues would drop by $25 million sequentially, while transaction costs would increase by $13 million. That would take the ratio of transaction costs as a percent of revenue from 49% just reported, to 57%. That would seem to be a forecast that would require much more of a mix swing than what the company has suggested. The company is forecasting OpEx of about $212 million, or a sequential increase of around 10%. That seems like a reasonable cadence, for a company with growth opportunities such as this one has, but of course, if revenues fall by 8% sequentially, as the company suggested, and OpEx rose by 10%, the ratio of OpEx to revenue is going to show disturbing trends. Specifically, the forecast would imply that OpEx as a percent of revenue would rise from 53% of revenues to 63% of revenues. My own model is based on a 58% ratio of OpEx to revenues to Q3, with improvements over the following 2 quarters.

Overall, I believe the revenue guide is far less than what the most likely result might be, and I don’t see such a violent increase in the transaction cost ratio, even given the mix change implied during the conference call.

How can Affirm grow so rapidly and why has it had such success acquiring merchant partners and retail customers?

Lost amongst all of the investor angst about revenue guidance and margin concerns has been the reasons for Affirm’s success. Just as it was true that Affirm’s valuation post its announcement of the Amazon partnership reached unsustainable levels, so now it seems as though the value of the Amazon partnership and Affirm’s other integrations is being totally ignored. Affirm has, and is likely to maintain a competitive moat that other competitors in the BN/PL space are unlikely to breach for the foreseeable future.

This period in the market has not been one in which investors have been paying much attention to the advantages of technology. And yet, it is technology that has enabled Affirm to be successful in establishing the relationships it has forged with Amazon, Walmart, Target, Shopify, American Airlines (AAL), Verifone and a host of other vendors. The following quote from the presentation of Affirm’s CEO on this latest call is no doubt a commercial, but it does suggest some of the material differentiation that Affirm maintains between the functionality it offers, and that of many competitors.

As a result, we avoid much of the adverse selection that often comes with traditional lending. Coupled with the SKU level data we receive from our partners our models tend to split the risk far better than those used in traditional consumer loans.

Another fundamental structural advantage Affirm has, is its total separability of transactions. Unlike providers of lines of credit, we underwrite transactions individually modeling a consumer’s ability to pay us back, as well as their propensity to do so. This notion of separability is also recursive a consequence of our product because repayment schedules are highly predictable our models operate at an individual installment level. This separability is a powerful tool for modeling as well as managing risk. We’re able to deliver a reliable forward-looking picture of both consumers and our own cash flow.

Our proprietary network of directly integrated merchants as well as other sources of non-traditional underwriting data offers us a significant raw data advantage into feature engineering. We maintain a library of over 500 features that we select from as we create new models or update existing ones, while continuously looking for and eliminating any potential for disparate impact in our decisioning both at individual variable and model levels. We train our models using academically well-understood [and] boosting technique with significant proprietary modifications we’ve invented that help us improve results. Because from the very beginning, we focused equally on consumer and merchant information, we ended up with a large number of models that are specific to our products and merchants who use them. Moreover, as we launch new products with new and existing partners, we acquire new types of data that we incorporate into the models and over time give incremental weight too.

Underwriting models decay over time, as macroeconomic conditions and consumer behaviors change. Even the very best performing ones can lose a few percentage points of their area under the curve every few months. Over the years, we’ve built special-purpose models that track model decay, the machine learning equivalent of a canary in the coal mine. Our proprietary software and processes allow us to rapidly retrain, retest and redeploy models where the performance has deteriorated in a matter of days

Making better underwriting decisions is perhaps the gold standard of building a long-term, high growth, profitable lending business. Affirm’s technology has allowed it to offer more flexible products that best suit the needs of its merchant partners and consumers. Unlike most of the many other BN/PL solutions that are in the space, Affirm has a much more flexible set of offerings. One of its latest offerings is called Adaptive Checkout. Adaptive Checkout offers far more choices to consumers than the standard offering in the space as consumers can choose payment lengths and other loan terms based on the characteristics of their purchase and their repayment capability. The results of using Adaptive Checkout are shown in the above link.

Last month, Affirm announced what it calls a Super App, which is a Chrome Browser extension which allows consumers to use Affirm, even in cases where Affirm is not listed as a payment option. A few months ago, the company launched its Debit+ card. The company hasn’t forecast the GMV or the revenue it expects from this offering. On the other hand, the CEO has said on several occasions that this offering will be a material source of future revenues. Most recently he commented as follows:

You may have noted the announcement of Visa being the launch partner for Debit+ which is now in the initial waitlist rollout. Though I will continue to caution you to not yet give crazy forecasting this product’s impact on our P&L at scale, I do have fun insight to share. On average, the number of weekly transactions by Debit+ consumer, excluding our own employees is greater than an order of magnitude above that of a regular Affirm user. Of course, these are enthusiastic early adopters and we fully expect the number to normalize, but it’s exciting to see first glimpses of what Affirm as a daily instrument might look like.

It is the kind of offering that is facilitated by Affirm’s core underwriting technology which provides the company with the ability to make much more extensive and flexible offers to consumers. I don’t imagine that there are too many wild and crazy forecasts regarding Affirm’s growth in the wake of the guidance most recently provided, and the consequent share price implosion. I do believe that long-term investors evaluating Affirm might choose to consider these new offerings as a tailwind of significance.

Investors these days are apparently not valuing anything they can’t touch and feel either in reported numbers or guidance, But I think it would be surprising if the new offerings and the new partnerships weren’t substantial tailwinds to revenue projections, both for this quarter and on into the foreseeable future. As mentioned, Affirm made substantial investments in both sales and marketing and in technology, and has rolled out a number of new services. The OpEx of doing so are what is visible now in the cost ratios. The revenues from those services should provide for hyper-growth into the foreseeable future.

Valuing Affirm/Wrapping Up

In the current environment in which investors are focused on both macro risks and immediate gratification in terms of earnings and cash flow, valuing Affirm can be a tendentious exercise. I am a long-term investor, for the most part, and while I do trade positions when I believe valuations have become excessive, for the most part, I try to maintain holdings for years, and have limited my trading to situations in which I perceive a change in the economic basis of a company. I acquired most of my AFRM holding at around $65/share 10 months ago, and sold some of it at $151 when it spiked in the wake of the announcement of the Amazon transaction.

I believe that after this latest share price implosion, the valuation of Affirm shares is unusually attractive. I have tried to suggest in this article the reasons why I find it necessary to construct my own estimates for forward revenue which are not necessarily congruent with either the company guidance, or the published First Call consensus. Using what I believe to be a conservative approach, I believe that the Affirm is likely to achieve revenue of about $1.85 billion over the next 4 quarters. At this point, the consensus forecast, as published by First Call, had a revenue outlook for the same period of $1.58 billion. I am expecting growth of 28% from the run rate the company reported last quarter; the consensus estimate is for growth of less than 10%. I think my own forecast is very conservative, and the consensus forecast fails any test I can think of with regards to reasonableness.

For the first 6 months of the year, Affirm’s free cash flow margin has been negative 12%. That’s noticeably worse than the 8.4% operating margin loss the company reported for the same period. There isn’t a whole lot of relevant history with regards to cash flow and its seasonality. What there is, shows that a significant component of the cash burn has come from negative balance sheet sources, primarily other assets and liabilities and payables.

As mentioned earlier, a forecast for elevated growth in the next several quarters is going to have material impacts on profitability and free cash flow. That said, I have still forecast cash burn at modest rates over the next 18 months. After that, however, the impact of the mix swing will likely reverse, which results in a rapid swing to much higher free cash flow margins. Of course, if Affirm’s growth rate continues at hyper levels of greater than 50%, the cross-over point will be deferred but the ramp will be far greater than I am willing to project. Overall, there appears to be a lot of leverage in the model which was evident last quarter.

After the implosion of the share price, and a positive revision to revenue growth, my derived EV/S ratio for the next 12 months has fallen to 6.5x. Of course, that is more than 50% below average for the company’s growth cohort. While its current free cash flow margin is below average, the impact of mix is not something that can be readily analyzed through the use of a single number.

After my revisions to the model, which raised revenues and increased cash burn over the coming 18 months, I now show Affirm shares to be valued at less than half the NPV I estimate. I have presented my assumptions; some readers will believe that I am too optimistic, particularly with regards to the trajectory of free cash flow margins. I can only suggest that the growth of interest income last quarter is a sign that over time, free cash flow margins will rival those of highly profitable software companies. I obviously have no way of “proving” that my assumptions are valid.

Catching falling knives is not considered good form in the investment world, and Affirm shares have been nothing other than falling knives. I imagine it will take patience before many analysts and investors emerge from hibernation and start to look at the details of the latest quarter and guidance and draw their own conclusions. In the meanwhile, I will start adding to my position in the name on a gradual basis, and expect strongly positive alpha from this base over the coming 12 months.