Why
A stablecoin card where the user pays in stablecoins is the obvious story and the boring one. The load-bearing move sits one layer up, on the issuer's balance sheet: a card program must fund Visa's daily net settlement before it has collected a cent from cardholders, so it needs short-term working capital every single day. Large issuers get that from banks; a new stablecoin-card program has no bank relationship and no track record, so its funding is either expensive or unavailable. That gap — not consumer checkout — is where the on-chain part earns its keep.
Two things make it worth a PoC rather than a headline. First, the underwriting collateral is behavioral, not asset: the credit limit is set from Visa's real settlement data — the issuer's observed throughput and repayment history — so the thing being lent against is a data stream, not posted BTC or cash. Second, repayment is structural, not promised: incoming receipts pass through a smart-contract waterfall that repays the draw before the issuer keeps the remainder, so 'will they pay it back' stops being a trust question and becomes a code path. Credit Coop reports 3,000+ draws and 9,000+ repayments on-chain with zero defaults — that zero is the waterfall's claim to make, and it's exactly the kind of claim a Jayverse PoC exists to test rather than repeat.
It also lands squarely on the umbrella thesis that settlement rails are infrastructure we USE, not build. Visa is the rail; the novel layer is financing the working capital that rail demands, on-chain, against data. The 'just-in-time settlement funding' variant is the sharpest version — borrow exactly the settlement shortfall at settlement time instead of holding a pre-borrowed buffer — and it is a clean thing to model: the cost you remove is idle borrowed capital, and the risk you take on is latency between the draw and the settlement deadline.
How it works
The daily gap the funding fills
A card program owes Visa on Visa's clock, but collects from cardholders on the cardholder's clock. The mismatch is a daily working-capital hole:
| Step | Timing | Who is short |
|---|---|---|
| Cardholder spends | T0 | — |
| Issuer must fund Visa net settlement | T0 (same day) | issuer |
| Cardholder repayment / prefunded balance settles | T+n | — |
The issuer draws stablecoin at T0 to cover the settlement, and repays from receipts at T+n. On-chain, both the draw and the repayment are transactions, and a contract can allocate the incoming receipts to repayment automatically.
What is actually novel — two things, not the stablecoin
- Underwriting on a data stream, not collateral. The credit limit is derived from Visa's real settlement data for that program — throughput and repayment history — so a program with no assets but a clean settlement record can still borrow. Karta reportedly started on a small revolving line and, after building an on-chain track record, secured a $125M institutional credit line. The track record is the collateral.
- Repayment as a waterfall, not a promise. Receipts flow through a smart contract that pays down principal + interest before releasing the remainder to the issuer. 'Zero defaults across 9,000+ repayments' is a property of that structure, not of borrower goodwill.
Just-in-time settlement funding
The sharper variant: instead of pre-borrowing a buffer and paying to hold idle capital, draw exactly the settlement shortfall at settlement time, sized from Visa's daily settlement data. The cost removed is idle-capital carry; the risk added is timing — the draw must clear before the settlement deadline. Visa reports up to ~30% lower funding cost for some programs using settlement data + on-chain history this way.
The Jayverse PoC
Model the loop on a local fork, not the numbers: a mock issuer with a daily settlement obligation, a revolving credit-line contract whose limit is a function of a fed-in 'settlement throughput' series, and a repayment waterfall on incoming receipts. Then test the invariants that actually matter — the credit line can never be drawn past its limit, the waterfall always repays before releasing, and (the just-in-time case) a draw that misses the settlement deadline is the failure mode to surface, not hide. Cross-refs: verex settlement (the receipts side), receipt-is-not-settlement (a payment intent is not a settled balance), and the 'rails we use not build' thesis.