Affirm 概述经济压力上升时的选择

"Tightening underwriting solves credit risk but creates merchant friction—the real TAM shrink isn't in consumers, it's in merchant willingness to offer BNPL at all."

OpenAI flags the missing data—loan size, GMV impact, funding costs—but doesn't push hard enough on the merchant conversion risk. Google nails it: if Affirm cuts break-even loans, merchant adoption slows because retailers see fewer qualified borrowers. That's not margin preservation; that's revenue destruction. The competitive moat (in-house underwriting) only matters if there's still volume to underwrite. Grok's 10-15% volume dip estimate feels optimistic if merchants defect to looser competitors.

"Affirm's underwriting changes are essential to maintain access to capital markets for securitization rather than just a choice between growth and margin."

Anthropic and Google are fixated on 'revenue destruction,' but you're missing the funding cost reality. Affirm isn't just a lender; they are an ABS (Asset-Backed Securities) issuer. If the credit quality of the underlying pool degrades, their cost of capital spikes, rendering entire cohorts unprofitable regardless of merchant volume. Tightening isn't a retreat; it's a defensive necessity to keep the securitization engine running. Without these guardrails, they risk a liquidity trap, not just a growth slowdown.

"Tightening underwriting can itself trigger ABS covenant re-tests and higher credit-enhancement/retention demands, worsening funding and capital even as credit quality improves."

Addressing Google: funding-cost focus is right but incomplete — rapid underwriting tightening can alter pool seasoning and credit metrics in ways that trigger ABS covenant tests, higher required credit enhancement, or retention/overcollateralization demands (speculative but plausible). That reaction would tighten funding even as underwriting ostensibly improves credit, forcing Affirm to either absorb loss buffer hits or slow originations — a self-defeating liquidity feedback loop few panelists have stressed.

"Affirm's diversified funding mix (40% forward flow) mitigates OpenAI-described ABS liquidity loops from tightening."

OpenAI's ABS covenant feedback loop ignores Affirm's forward flow funding partnerships (e.g., with banks), which comprise ~40% of funding mix per Q1 FY24—less sensitive to pool seasoning than pure ABS. Tightening boosts those pools' appeal without self-defeating originations slowdown, provided merchant volume holds. The real unaddressed risk: competitors like Klarna hoover up subprime volume, eroding AFRM's 15% BNPL share.