AFRM · Forward model · Levchin case
The Levchin case, 20 quarters out
Each segment is projected from its own operating driver, rolled up into consolidated cash flow, and discounted back to a fair value per share. The assumptions below are editable — change them and every number on this page moves with them.
Affirm reports five revenue lines on the face of every shareholder letter - merchant network, card network, interest income, gain on sales of loans, servicing income - and this model carries all five exactly as filed for six quarters, from FQ2'25 (December 2024) to FQ3'26 (March 2026). Nothing is apportioned and no line is invented, so no quarter is flagged estimated. Quarter labels are FISCAL: '2026 Q3' is the quarter ended 31 March 2026, because Affirm's year ends 30 June. What IS assumed is the cost side. Affirm publishes no segment economics at all, so the split of the four transaction-cost lines across the five revenue lines is ours: funding costs and the provision for credit losses are charged entirely to interest income, the loss on loan purchase commitment entirely to merchant network (it arises on the 0% promotional loans merchants pay for), and processing and servicing is split across the four remaining lines. The five margins are set so they sum to the reported $498M of RLTC in FQ3'26, and corporate overhead at 21% of revenue then reproduces the reported 27.0% adjusted operating margin exactly. Two known distortions, stated rather than tuned away. First, the engine has no seasonality: merchant network and gain on sale grow at a smooth sequential rate, so the guided June quarter ($13.15-13.45B of GMV) and every December quarter come out light while March quarters come out heavy, and fiscal-year sums grow more slowly than the quarterly run rate does. Second, corporate overhead is a fixed share of revenue here, so the operating leverage Affirm promises - the same RLTC yield carrying a lower opex ratio - has to be carried in each vertical's terminal margin instead of in a falling overhead line.
Affirm's own medium-term framework taken at face value: 25%+ GMV growth to $100B a year, revenue at 7.5-8.5% of it, RLTC at 3.75-4.0%, and a 30-35% adjusted operating margin. This case reaches the framework early rather than differently - $100B of annual GMV lands around 2029 instead of 2030, and margins settle at the top of the published band. What it does NOT reach is the version of the story where the Card and the adjacent verticals make Affirm something other than a lender: even here, interest income plus gain on sale is still more than half of terminal revenue, and the merchant fee rate still erodes. The framework is a bigger Affirm, not a different one - and that gap, not the growth rate, is what the multiple is arguing about. Which is why this case carries a 6x exit rather than the bull case's 6.5x: the company's own numbers, delivered in full, still describe a lender.
Latest: $3.47B (2031Q3E)
| Period | Value |
|---|---|
| 2025Q2 | $866M |
| 2025Q3 | $783M |
| 2025Q4 | $876M |
| 2026Q1 | $933M |
| 2026Q2 | $1.12B |
| 2026Q3 | $1.04B |
| 2026Q4E | $1.12B |
| 2027Q1E | $1.21B |
| 2027Q2E | $1.30B |
| 2027Q3E | $1.40B |
| 2027Q4E | $1.50B |
| 2028Q1E | $1.61B |
| 2028Q2E | $1.72B |
| 2028Q3E | $1.84B |
| 2028Q4E | $1.96B |
| 2029Q1E | $2.10B |
| 2029Q2E | $2.24B |
| 2029Q3E | $2.38B |
| 2029Q4E | $2.54B |
| 2030Q1E | $2.70B |
| 2030Q2E | $2.88B |
| 2030Q3E | $2.99B |
| 2030Q4E | $3.11B |
| 2031Q1E | $3.22B |
| 2031Q2E | $3.34B |
| 2031Q3E | $3.47B |
What drives each segment
Merchant network revenue
Units × priceThe fee merchants pay Affirm for the sale, and the closest thing this company has to a pure take rate on volume. It is highest on 0% promotional and Pay-in-X plans, where the merchant is buying the conversion rather than the consumer buying the credit, so this line rises and falls with how much of the marketplace is merchant-funded rather than with GMV alone.
Latest: $784M (2031Q3E)
| Period | Value |
|---|---|
| 2025Q2 | $245M |
| 2025Q3 | $214M |
| 2025Q4 | $239M |
| 2026Q1 | $251M |
| 2026Q2 | $328M |
| 2026Q3 | $268M |
| 2026Q4E | $286M |
| 2027Q1E | $306M |
| 2027Q2E | $327M |
| 2027Q3E | $350M |
| 2027Q4E | $374M |
| 2028Q1E | $401M |
| 2028Q2E | $429M |
| 2028Q3E | $460M |
| 2028Q4E | $493M |
| 2029Q1E | $528M |
| 2029Q2E | $566M |
| 2029Q3E | $607M |
| 2029Q4E | $651M |
| 2030Q1E | $699M |
| 2030Q2E | $750M |
| 2030Q3E | $760M |
| 2030Q4E | $765M |
| 2031Q1E | $771M |
| 2031Q2E | $777M |
| 2031Q3E | $784M |
Assumptions & reasoning
- The unit here is one million dollars of GMV, so this line is Affirm's volume multiplied by the merchant fee it earns per million. Both halves are disclosed: GMV every quarter in the headline, the fee derived by division, and the resulting 2.31% has held between 2.30% and 2.49% across all six reported quarters.
- GMV grows 6.5% a quarter here, which is a 28.6% annual run rate against the 25%-plus the company put in its medium-term framework. That is deliberately above the framework and below the 35% Affirm actually delivered in the March quarter.
- The ceiling of $30B of GMV a quarter - $120B a year - is what stops a 6.5% compound running to infinity, and it binds from about the fourth projected year. Affirm says $28.0B of committed funding capacity supports more than $65B of annual GMV, so this ceiling assumes that capacity is roughly doubled again.
- The fee rate drifts down 0.75% a quarter toward a much slower long-run decline. Volume moving to the Affirm Card and to wallet partners is volume where no merchant integration is paying a network fee, and that mix shift is the single clearest threat to this line.
- This is the same GMV that drives the gain-on-sale line, and the two carry it separately: moving volume here without moving it there breaks the arithmetic that ties both to one company.
Card network revenue
Subscribers × ARPUInterchange and virtual-card economics on the direct-to-consumer surface: the Affirm Card in the wallet and at the terminal, plus the virtual cards that carry wallet-partner volume. This is the engine Affirm's own framework leans on hardest - more than ten percentage points of annual growth from D2C - and it is the one line where the driver is a person rather than a purchase.
Latest: $305M (2031Q3E)
| Period | Value |
|---|---|
| 2025Q2 | $58M |
| 2025Q3 | $59M |
| 2025Q4 | $67M |
| 2026Q1 | $69M |
| 2026Q2 | $73M |
| 2026Q3 | $66M |
| 2026Q4E | $76M |
| 2027Q1E | $85M |
| 2027Q2E | $95M |
| 2027Q3E | $105M |
| 2027Q4E | $116M |
| 2028Q1E | $126M |
| 2028Q2E | $137M |
| 2028Q3E | $148M |
| 2028Q4E | $159M |
| 2029Q1E | $171M |
| 2029Q2E | $183M |
| 2029Q3E | $195M |
| 2029Q4E | $208M |
| 2030Q1E | $220M |
| 2030Q2E | $234M |
| 2030Q3E | $247M |
| 2030Q4E | $261M |
| 2031Q1E | $275M |
| 2031Q2E | $290M |
| 2031Q3E | $305M |
Assumptions & reasoning
- Modelled as active cardholders times revenue per cardholder because those are the disclosed numbers that move this line: 4.4 million active cardholders and $2.1B of Card GMV in FQ3'26, with in-store Card GMV up 170%.
- The $5.03 a month is derived, not disclosed, and it is doing more work than a per-cardholder figure should. Card network revenue also carries virtual-card economics on wallet-partner volume that no cardholder generates, so read this as revenue per cardholder-equivalent rather than as what a cardholder is worth.
- The attach frame is the honest part: 4.4 million cardholders against 26.8 million active consumers is 16.4% attach, and the ceiling of 55% says a majority of Affirm's own consumers eventually carry the card - aggressive, approached asymptotically, and never reached inside this horizon.
- Attach reaches about 40% of a consumer base growing 2% a quarter by 2031, which triples this line. Affirm Card GMV grew 159% in the December quarter and cardholders 121%, so the model decelerates that hard rather than extrapolating it.
- Revenue per cardholder drifts down 0.5% a quarter with a floor at $3.00. Card 0% APR GMV is already nearly 20% of the Card mix and rising, and 0% volume earns a merchant fee rather than interchange.
Interest income
Capacity × utilisation × priceThe largest line in the company and the one that is genuinely supply-gated: Affirm can only earn interest on loans it funds and keeps. Loans held for investment were $8.57B at 31 March 2026 against $28.0B of committed funding capacity, so the constraint on this line is not demand for credit but how much of it Affirm chooses - and is able - to hold.
Latest: $1.83B (2031Q3E)
| Period | Value |
|---|---|
| 2025Q2 | $409M |
| 2025Q3 | $403M |
| 2025Q4 | $419M |
| 2026Q1 | $454M |
| 2026Q2 | $494M |
| 2026Q3 | $532M |
| 2026Q4E | $575M |
| 2027Q1E | $619M |
| 2027Q2E | $665M |
| 2027Q3E | $713M |
| 2027Q4E | $763M |
| 2028Q1E | $815M |
| 2028Q2E | $870M |
| 2028Q3E | $927M |
| 2028Q4E | $986M |
| 2029Q1E | $1.05B |
| 2029Q2E | $1.11B |
| 2029Q3E | $1.18B |
| 2029Q4E | $1.25B |
| 2030Q1E | $1.32B |
| 2030Q2E | $1.40B |
| 2030Q3E | $1.48B |
| 2030Q4E | $1.56B |
| 2031Q1E | $1.65B |
| 2031Q2E | $1.74B |
| 2031Q3E | $1.83B |
Assumptions & reasoning
- Capacity here is the loan book itself: $8,573M of loans held for investment at 31 March 2026, straight off the balance sheet. Utilisation is the share of that book that actually carries a rate, because the 0% promotional loans sitting in it earn a merchant fee instead.
- The $73,940 of quarterly revenue per $1M of interest-bearing balance is a 7.4% quarterly yield, near 30% annualised. That looks extreme until you remember these are five-month loans with APRs up to 36% and that interest income here includes the accretion of discounts on retained paper.
- The book grows $600M in the first projected quarter and the addition itself grows 3% a quarter, which decelerates the percentage growth of the balance as it compounds. Loans held for investment grew from $7.03B to $8.57B in the nine months to March, so $600M a quarter is the recent pace, not an acceleration of it.
- Yield drifts down 0.5% a quarter toward a much slower decline. Affirm priced its third consecutive revolving ABS under a 5% blended yield in March and reports its lowest cost of funds in three and a half years - cheaper funding is competed away in consumer APR before it is kept.
- This is the line that carries the credit cycle. The provision for credit losses and funding costs are both charged here in full, which is why its margin sits near 42% rather than near the 90% a gross interest line would show, and why a bear case has to move this margin more than any other.
Gain on sales of loans
Units × priceWhat Affirm books when it sells originated paper to forward-flow buyers and securitisation trusts instead of holding it. It is the release valve on the balance sheet: every loan sold is capital freed to originate again, and the price it fetches is the market's live opinion of Affirm's underwriting.
Latest: $401M (2031Q3E)
| Period | Value |
|---|---|
| 2025Q2 | $125M |
| 2025Q3 | $76M |
| 2025Q4 | $117M |
| 2026Q1 | $119M |
| 2026Q2 | $185M |
| 2026Q3 | $127M |
| 2026Q4E | $137M |
| 2027Q1E | $147M |
| 2027Q2E | $158M |
| 2027Q3E | $170M |
| 2027Q4E | $183M |
| 2028Q1E | $197M |
| 2028Q2E | $212M |
| 2028Q3E | $228M |
| 2028Q4E | $245M |
| 2029Q1E | $264M |
| 2029Q2E | $284M |
| 2029Q3E | $305M |
| 2029Q4E | $328M |
| 2030Q1E | $353M |
| 2030Q2E | $380M |
| 2030Q3E | $386M |
| 2030Q4E | $390M |
| 2031Q1E | $394M |
| 2031Q2E | $397M |
| 2031Q3E | $401M |
Assumptions & reasoning
- Affirm does not disclose how much paper it sells in a quarter or at what price, so this line is modelled on the denominator it does disclose: GMV. The unit is one million dollars of volume, and the $10,966 it earns per million is a derived rate, not a published one.
- That rate has run between $8,814 and $13,422 per $1M of GMV across the six reported quarters with no trend, which is why the drift is set to zero. It moves with how much Affirm chooses to sell and what buyers will pay, and neither is forecastable from outside.
- GMV growth and the ceiling here are identical to the merchant network line by construction, because it is the same volume. If you move one, move the other, or the model is describing two different companies.
- The margin near 91% is the highest in the model and the most fragile assumption in it: the credit cost of sold loans sits with the buyer, so almost nothing lands against this revenue except a share of processing. A forward-flow market that reprices does not lower this margin, it removes the revenue.
- This is the line that goes first in a funding shock. In 2022 the securitisation market closed to consumer paper for months, and a company with $28.0B of committed capacity still needs a bid for what it originates beyond what it can hold.
Servicing income
Growth pathThe fee Affirm earns for collecting on loans other people own. It is the annuity attached to the sold book - small, boringly consistent, and the one line that keeps paying after Affirm has stopped taking the credit risk.
Latest: $152M (2031Q3E)
| Period | Value |
|---|---|
| 2025Q2 | $29M |
| 2025Q3 | $32M |
| 2025Q4 | $34M |
| 2026Q1 | $40M |
| 2026Q2 | $43M |
| 2026Q3 | $45M |
| 2026Q4E | $48M |
| 2027Q1E | $52M |
| 2027Q2E | $55M |
| 2027Q3E | $59M |
| 2027Q4E | $63M |
| 2028Q1E | $67M |
| 2028Q2E | $72M |
| 2028Q3E | $76M |
| 2028Q4E | $81M |
| 2029Q1E | $86M |
| 2029Q2E | $92M |
| 2029Q3E | $97M |
| 2029Q4E | $103M |
| 2030Q1E | $109M |
| 2030Q2E | $115M |
| 2030Q3E | $122M |
| 2030Q4E | $129M |
| 2031Q1E | $136M |
| 2031Q2E | $144M |
| 2031Q3E | $152M |
Assumptions & reasoning
- This is the one line in the model with no driver behind it, and that is the honest answer rather than a lazy one: the thing that decides servicing income is the balance of third-party-owned loans Affirm services, and Affirm does not publish that balance quarterly.
- So it runs on a growth rate instead, starting at 6.5% a quarter and decaying toward 4%. The six reported quarters compounded at 7.6% a quarter and grew 39% year over year, so the opening rate is already a step down from the record.
- At 4.3% of revenue this is the smallest line in the model, and a wrong growth rate here changes the fair value by cents. It earns its place because leaving it out would break the reconciliation to reported total revenue.
- The margin near 55% is a guess constrained by arithmetic: it is what is left after a share of processing and servicing costs, and it has to sit where the five lines together sum to the $498M of RLTC Affirm reported for the quarter.
- The direction of risk is upward. Every quarter Affirm sells more paper than it holds, the serviced book that pays this fee gets larger - which is why a bear case for gain on sale is not automatically a bear case here.
Where each case comes from
Levchin case — primary sources
The primary sources this case is built from — filings, calls and posts. Where they are claims by an interested party rather than disclosures, every number in the Levchin column is what happens if they are taken at face value.
The growth the framework is indexed to
- May 7, 2026 We delivered 35% GMV growth, our 10th consecutive quarter of over-30% growth – without compromising unit economic discipline. Revenue grew 33% and RLTC 41%.
- May 7, 2026 While top-line growth continues at a very brisk pace, Affirm is maturing like fine wine further down the P&L.
- Feb 5, 2026 We estimate that Affirm grew more than 5x the growth rate of overall U.S. credit card spend in 2025 and 4x the rate of e-commerce growth.
What the funding side has to deliver
- May 7, 2026 We estimate this capacity can support more than $65.0 billion in annual GMV based upon our weighted-average loan duration of approximately 5 months at the end of FQ3'26.
- May 7, 2026 We also priced our third consecutive revolving ABS transaction under a 5% blended yield in March, and our average annualized cost of funds reached the lowest level in three and a half years.
The direct-to-consumer engine
- Feb 5, 2026 Affirm Card also had a banner quarter, growing GMV 159%, and active cardholders rose 121%.
- Feb 5, 2026 Card 0% APR GMV increased 190% year over year, now representing almost 20% of the total Card GMV mix, a 4 percentage point increase sequentially, a nearly threefold increase in mix over two years.
From cash flow to fair value
| Present value of free cash flow, 20 quarters | $7.53B |
| Terminal-year revenue | $13.14B |
| Terminal-year EBITDA | $4.72B |
| Exit multiple, on revenue | 6.0x |
| Terminal value | $78.87B |
| Discounted at 12.0% a year, terminal value becomes | $44.75B |
| Enterprise value | $52.28B |
| Net cash | $1.35B |
| Equity value | $53.63B |
| Diluted shares | 0.35B |
| Fair value per share | $153.50 |
| Against the current price of $75.05 | +105% |
5x terminal revenue against roughly 6.3x trailing at $77.33 today. A business still compounding near 20% with a 33% adjusted operating margin deserves a real multiple, but it is a levered consumer lender at the end of this horizon rather than a hypergrowth network, and paying today's multiple for that is an argument this model does not make. The 12% discount rate is deliberately above what a software business would carry: Affirm funds a loan book, and the cost of that funding is the risk.
Read the other way round: at $75.05 the market is paying 2.3x terminal-year revenue, holding every other assumption on this page fixed. That is the number to argue about.
The projected path
| Quarter | Merchant network revenue | Card network revenue | Interest income | Gain on sales of loans | Servicing income | Revenue | YoY | EBITDA | Capex | FCF | R40 | PV of FCF |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2026 Q4E | $286M | $76M | $575M | $137M | $48M | $1.12B | +28% | $364M | $63M | $235M | +49 | $228M |
| 2027 Q1E | $306M | $85M | $619M | $147M | $52M | $1.21B | +30% | $397M | $66M | $258M | +51 | $244M |
| 2027 Q2E | $327M | $95M | $665M | $158M | $55M | $1.30B | +16% | $432M | $69M | $283M | +38 | $260M |
| 2027 Q3E | $350M | $105M | $713M | $170M | $59M | $1.40B | +35% | $468M | $73M | $308M | +57 | $275M |
| 2027 Q4E | $374M | $116M | $763M | $183M | $63M | $1.50B | +34% | $507M | $76M | $336M | +56 | $291M |
| 2028 Q1E | $401M | $126M | $815M | $197M | $67M | $1.61B | +33% | $548M | $80M | $365M | +56 | $308M |
| 2028 Q2E | $429M | $137M | $870M | $212M | $72M | $1.72B | +32% | $591M | $84M | $395M | +55 | $324M |
| 2028 Q3E | $460M | $148M | $927M | $228M | $76M | $1.84B | +32% | $637M | $89M | $427M | +55 | $341M |
| 2028 Q4E | $493M | $159M | $986M | $245M | $81M | $1.96B | +31% | $685M | $93M | $461M | +55 | $358M |
| 2029 Q1E | $528M | $171M | $1.05B | $264M | $86M | $2.10B | +31% | $736M | $98M | $497M | +54 | $375M |
| 2029 Q2E | $566M | $183M | $1.11B | $284M | $92M | $2.24B | +30% | $790M | $103M | $535M | +54 | $392M |
| 2029 Q3E | $607M | $195M | $1.18B | $305M | $97M | $2.38B | +30% | $847M | $109M | $575M | +54 | $410M |
| 2029 Q4E | $651M | $208M | $1.25B | $328M | $103M | $2.54B | +29% | $907M | $115M | $618M | +54 | $427M |
| 2030 Q1E | $699M | $220M | $1.32B | $353M | $109M | $2.70B | +29% | $970M | $121M | $663M | +53 | $446M |
| 2030 Q2E | $750M | $234M | $1.40B | $380M | $115M | $2.88B | +29% | $1.04B | $127M | $710M | +53 | $464M |
| 2030 Q3E | $760M | $247M | $1.48B | $386M | $122M | $2.99B | +26% | $1.08B | $131M | $739M | +50 | $469M |
| 2030 Q4E | $765M | $261M | $1.56B | $390M | $129M | $3.11B | +22% | $1.12B | $136M | $766M | +47 | $473M |
| 2031 Q1E | $771M | $275M | $1.65B | $394M | $136M | $3.22B | +19% | $1.16B | $140M | $794M | +44 | $477M |
| 2031 Q2E | $777M | $290M | $1.74B | $397M | $144M | $3.34B | +16% | $1.20B | $144M | $823M | +41 | $480M |
| 2031 Q3E | $784M | $305M | $1.83B | $401M | $152M | $3.47B | +16% | $1.24B | $149M | $853M | +41 | $484M |
Every row is projected. A year-over-year change is shown only where the quarter it compares against exists — an em dash means there is no comparable quarter, not a flat year.
Model revisions
Assumptions are marked to reality as each quarter prints. Every change is appended here, with the fair value the model produced at the time, so the model's own history stays visible.
| Date | Changed | Fair value then | Note |
|---|---|---|---|
| 2026-08-21 | all | $107.49 | First publication. Five reported revenue lines from FQ2'25 to FQ3'26, basis FQ3'26. Segment margins calibrated so they sum to the reported $498M of RLTC and 21% corporate overhead reproduces the reported 27.0% adjusted operating margin. Base fair value $107.49 against $77.33; bear $41.44, Levchin $153.50, bull $175.92. |