The Consumer Credit Ecosystem: From Rejected Application to Retirement Fund

The arrows point from origination to destination. Capital flows in the same direction. Credit risk follows the capital but takes longer to settle and is the subject of this report. The diagram compresses for readability: in practice a single loan can pass through Stages 3 and 4 multiple times via re-securitisation before arriving at Stage 5, and not every corridor uses every stage. The Pagaya step is specific to second-look approvals at SoFi and similar partner originators; direct origination flows skip it. Modeled

The originators are the public face of the system. They are the brand the consumer sees, the app the consumer downloads, the underwriting decision the consumer experiences. They are also, increasingly, the surface layer on a balance sheet that does not actually retain the credit risk they generate. Below are the four largest US-and-Europe-facing originators by relevant 2025-2026 metrics. All figures are sourced from earnings reports, investor presentations, or counterparty press releases.

Two pieces of public data anchor this section. First, the FHFA's published Credit Risk Transfer dataset shows that between 2013 and 2023 the GSEs transferred $6.7tn in unpaid principal balance and $210.2bn in Risk-in-Force to private investors. Second, Freddie Mac's ACIS reinsurance programme currently maintains a 50% concentration in its top 5 counterparties. The Federal Reserve's 2024 approval of six megabanks to issue their own Credit-Linked Notes (referenced in Fed SR 23-11) means the same concentration shape is now appearing in bank capital relief trades alongside the GSE programmes.

The Crassus reading of these data points is that the system has converged on a small handful of counterparties to absorb the credit-default risk of trillions of dollars of consumer and mortgage loans. We use the term concentration risk to describe this convergence.

In July 2026 the UK's Financial Conduct Authority brings Buy Now Pay Later providers under formal Consumer Duty and affordability assessment regulation. The implementation date is confirmed in FCA published materials. What is not directly confirmed by the FCA is the magnitude of the resulting acceptance-rate impact on the existing BNPL user base. The 30% acceptance-rate cut-off figure that has circulated in industry commentary is not, as best we have been able to verify, derived from any FCA-published BNPL-specific impact assessment. It is derived from a different regulatory event.

The closest analogue is the FCA's earlier crackdown on High-Cost Short-Term Credit (HCSTC), the regime change that effectively ended the payday-lending market in its 2014-2016 form. UK Parliament research and TransUnion industry analysis document that HCSTC acceptance rates fell from approximately 50% to approximately 30% under the affordability-check regime. This is the source of the "30%" number. We are calculating from the historical precedent, not from a direct FCA BNPL forecast.

This report maps an observable pipeline using public data. It does not predict the timing or magnitude of failure. It does not recommend trades, positions, or allocations. The contribution is the consolidation of named exposures into a single legible diagram and the acknowledgement that the same five firms recur at the absorber stage of every major fintech corridor.

Several limitations require explicit acknowledgement. First, the concentration risk framing in Section 7 is the original analytical contribution of this report and diverges from the rating agency consensus (Moody's January 2026 Private Credit Outlook). Where we and the rating agencies look at the same data and reach different conclusions, we have stated this explicitly. We are not asserting that Moody's is wrong; we are asserting that "growth acceleration built on five counterparties absorbing risk from every major originator simultaneously is indistinguishable from concentration risk until one of the five fails." The reader may agree, disagree, or remain undecided. The disagreement is the position.

Second, the structural observation in Section 5 about KKR's vertical integration with Global Atlantic is a Modeled analytical observation. No specific Fitch report on this structure exists in the public record as of publication. The structure itself is documented in KKR's Q4 2025 SEC filing. The implication that vertical integration of this kind carries elevated structural risk is ours.

Third, the 30% BNPL acceptance-rate cut-off figure used in Section 8 is derived from the FCA's earlier HCSTC regulatory precedent (TransUnion / UK Parliament), not from a direct FCA BNPL impact assessment. The directional inference is reasonable; the magnitude is an analogue.

Fourth, the CalPERS / OMERS / BCI / Kuvare transaction is confirmed by two independent public sources (Seeking Alpha / Bloomberg and CRISIL / Integral IQ). Whether the trade represents a standard pension allocation or a liquidity backstop for a private credit manager facing redemption pressure is a matter of interpretation that the public record does not settle. We have presented both readings in Section 6.