Most people see a $10 billion revenue milestone and think 'growth story.' I see a balance sheet stress test that hasn't been graded yet. Affirm just reported revenue exceeding $10 billion and raised guidance. BNPL momentum is accelerating, they say. Market expansion is underway. Fine. But I didn't get here by reading press releases. I got here by auditing the mechanics underneath the hype.
Let's cut through the noise. The market is treating this earnings beat as a validation of the buy-now-pay-later thesis. It's not that simple. Revenue growth without unit economics is just a more expensive way to lose money. The question isn't whether Affirm can generate demand. It's whether that demand is profitable at scale. And the answer, based on what the earnings report didn't say, is far less certain than the stock price suggests.
The Context: What Affirm Actually Is
Affirm is not a payment company. It's a consumer credit company wearing a fintech costume. The core product is installment lending at point-of-sale, primarily in e-commerce. The business model is straightforward: merchants pay a discount fee for higher conversion rates, and consumers pay interest on longer-term installment products. The technology differentiator is real-time credit decisioning. A machine learning model evaluates a borrower's creditworthiness in roughly one second using non-traditional data points.
This matters because Affirm's target demographic is younger consumers with thin credit files. These are people traditional banks won't touch. Affirm's underwriting model claims to assess this risk better than legacy FICO-based systems. That claim is the entire investment thesis. If the model works, Affirm unlocks a massive underserved market. If the model fails during a downturn, the losses compound quickly.
The regulatory architecture is equally important. Affirm operates through a bank partnership model. Loans are originated by partner banks like Cross River Bank, not by Affirm itself. This structure allows Affirm to avoid holding a banking license while still offering credit products. It's legally efficient. It's also a point of regulatory vulnerability. If regulators decide the bank partnership model is a compliance dodge, the entire operating framework comes under pressure. I've seen this playbook before in crypto lending. It doesn't end well when the music stops.
The Core: What the Numbers Actually Tell Us
Revenue crossing $10 billion is not nothing. It validates that a genuine market exists for transparent, installment-based consumer credit. But revenue is a top-line vanity metric. The real indicators live in net loss rates, funding costs, and merchant concentration. The earnings report didn't disclose these numbers. That omission is a signal.
Let me break down the unit economics. Affirm makes money three ways: merchant discount fees, consumer interest income, and other services. The merchant fee is the largest driver. Merchants pay a premium because BNPL increases conversion rates, especially for high-ticket items. This creates a dependency. If merchants cut marketing budgets during an economic slowdown, Affirm's revenue takes a direct hit.
The interest income side is more concerning. Affirm offers both zero-interest and interest-bearing products. Zero-interest products are subsidized by merchant fees. Interest-bearing products expose Affirm to credit risk. In a rising-rate environment, funding costs increase. If the consumer loan portfolio grows faster than the merchant fee base, margins compress. The earnings report doesn't tell us which segment drove the revenue growth. That distinction matters enormously.
I spent six weeks in 2020 writing Python scripts to arbitrage yield farming inefficiencies. I learned that gross returns mean nothing until you subtract gas costs, impermanent loss, and slippage. The same logic applies here. Gross revenue means nothing until you subtract funding costs, default losses, and operational expenses. Affirm's adjusted operating income improved. But adjusted metrics in fintech often exclude the exact line items that kill you later.
The risk concentration is the real story. Affirm's partnership with Amazon is a significant revenue source. I don't have the exact percentage from the report, but industry observers estimate it's substantial. This is a single-point-of-failure risk. If Amazon builds its own BNPL product or renegotiates terms unfavorably, Affirm's top line takes a step function down. No amount of technology moat protects you from a counterparty with more leverage than you have. Amazon doesn't need Affirm. Affirm needs Amazon. That's an uncomfortable position.
The Contrarian Angle: The Real Threat Isn't Bad Debt. It's Dependency.
Everyone focuses on credit risk when discussing BNPL. Defaults are the obvious danger. But I'd argue the more insidious risk is structural dependency. Affirm's model depends on three external parties: bank partners for origination, large merchants for distribution, and capital markets for funding. Each dependency is a potential failure point.
Bank partners face their own regulatory pressures. If Cross River or another partner scales back BNPL exposure, Affirm needs to find alternative origination channels. That's not a quick process. Merchant dependencies are even more fragile. Big-box retailers have negotiating power. They can demand lower fees or build in-house alternatives. The platform giants like Apple are already entering the space with Apple Pay Later. Apple has the user base and the ecosystem lock-in. Independent BNPL players are now competing against the most valuable company on earth. That's not a fair fight.
Then there's the funding side. Affirm relies on asset-backed securities and warehouse lines to finance its loan book. In a credit crunch, these funding sources evaporate. I lived through the 2022 bear market in crypto. I watched projects with real revenue die because they couldn't refinance. The mechanics are the same in traditional finance. When liquidity dries up, solvent businesses can still fail if their liabilities are short-term and their assets are illiquid. Affirm's loan book is the asset. The ABS market is the liability. If that market freezes, the model breaks.
Hype is a liability; liquidity is the only truth. The market is celebrating revenue growth while ignoring the liquidity architecture underneath. That's backwards.
The Takeaway: What I'm Watching Next
The path forward for Affirm is a narrow one. A potential Fed rate-cutting cycle would reduce funding costs and expand net interest margins. That's a genuine tailwind. But it only helps if credit losses remain contained. The current consumer environment is showing cracks. Delinquencies across the consumer credit spectrum are rising. Affirm's borrower base is riskier than the average bank's. They will feel the pain first.
I'm watching three signals. First, the net loss rate in the next quarterly report. If it jumps more than 200 basis points sequentially, that's a red flag. Second, any news about Amazon contract renewals. The timing and terms will tell us who holds the power. Third, the growth of Affirm Card active users. This is the key to converting transactional users into recurring account relationships. If that metric grows above 20% quarter-over-quarter, the strategy is working.
We do not predict the storm; we build the ship. Affirm has built a decent ship. But it's sailing in increasingly crowded waters with a hull that hasn't been tested in a real downturn. The revenue growth is real. The sustainability is unproven.
Trust the code, verify the chain, own the outcome. In this case, the code is the underwriting model. The chain is the partnership network. The outcome is whether this business generates durable profits or just delayed losses. I'm not short the stock. But I'm not buying the narrative either. I'm waiting for the data that actually matters. The next earnings report will tell us more than this one did. Until then, the disciplined position is observation. Not conviction.
Most people will chase the momentum. That's fine. The market needs liquidity providers. But I didn't build a trading community by following headlines. I built it by finding the gap between perception and reality. Right now, that gap is wide enough to drive a truck through. And I plan to be on the right side of it when it closes.