New payment rails are winning the flows where money moves between institutions and across borders in size, losing the flows where a payment is bundled with credit, disputes, and acceptance, and fighting the incumbents to a draw everywhere in between, which makes the real contest one over where margin sits rather than whether rails get replaced. That single sentence is the chapter's conclusion, and the 5 parts that follow earn it job by job. Most commentary on stablecoins and tokenised money fails before it starts by treating "payments" as one market with one incumbent; there are at least 5 distinct jobs, each with different incumbents, different unit economics, and different displacement odds.
The 5 jobs
- Point-of-sale consumer payments. The incumbent is the card networks' four-party model plus, in Asia, domestic instant-payment schemes and wallet super-apps. The buyer is a consumer who wants credit, rewards, and the right to dispute.
- Cross-border B2B payments. The incumbent is correspondent banking: chains of nostro accounts, cut-off windows, and multi-day settlement. The buyer is a corporate treasurer moving 5 to 7 figures who cares about certainty, cost, and working-capital float.
- Treasury and intercompany flows. The incumbent is the bank's own wires and sweep infrastructure. The buyer is the same treasurer moving money between the company's own entities, where nobody needs consumer protections at all.
- Remittances. The incumbents are money-transfer operators and, increasingly, linked domestic instant-payment schemes. The buyer sends hundreds of dollars home and pays the highest percentage costs in payments.
- Capital markets settlement. The incumbent is the CSD (central securities depository) and payment-system stack settling against central-bank or commercial-bank money. The buyer is an institution that cares about atomicity and collateral mobility, covered in depth in the atomic DvP (delivery-versus-payment) chapter.
The map
| Job | Verdict | Why | What would change it |
|---|---|---|---|
| Point-of-sale consumer | Incumbents hold | Cards are a bundle (credit, fraud liability, disputes, acceptance), and a bare transfer replicates none of it; domestic A2A (account-to-account) schemes already deliver cheap instant payment where the bundle is not wanted | Interchange regulation compressing the bundle's funding, or wallet-level stablecoin spending reaching acceptance parity |
| Cross-border B2B | New rails win share | Correspondent banking is the weakest incumbent: slow, opaque, and expensive at exactly the ticket sizes where ramp costs amortise; tokenised deposit networks are already settling material volume | Swift's shared ledger reaching production scale would let incumbents absorb the gain |
| Treasury and intercompany | Tokenised deposits win | The bundle is dead weight, counterparties are known, and bank-issued tokenised money keeps flows inside the regulatory perimeter; this is the quietest and most complete displacement under way | Little; this verdict has the most production evidence behind it |
| Remittances | Contested, corridor by corridor | Stablecoins win corridors that lack scheme links today; linked A2A schemes and Nexus-class multilateral plumbing are the structural competitor and are regulator-sponsored | Nexus going live at scale would close the window in ASEAN corridors; ramp-cost compression would extend stablecoin reach |
| Capital markets settlement | Hybridise | The cash leg tokenises (deposits, and MMF (money-market fund) shares as collateral) while the asset leg stays inside regulated wrappers; nobody displaces the CSD, the CSD adopts the rail | Already happening; the open question is which form of tokenised cash wins the settlement asset seat |
Part 2 takes cards seriously on their own economics, because the displacement case is only honest once the bundle is priced. Part 3 audits the stablecoin cost claim end to end. Part 4 gives the domestic instant-payment schemes their due as the quiet incumbents in Asia. Part 5 separates the 3 kinds of tokenised money that commentary blurs together. Part 6 reads the incumbent response and closes the margin-migration argument.
The vocabulary, in one pass
- Four-party model. The card structure linking cardholder, issuer (the cardholder's bank), merchant, and acquirer (the merchant's bank), with the network coordinating in the middle.
- Interchange. The fee the acquirer pays the issuer on each card transaction, set by network rules; it funds credit risk, rewards, and fraud liability, and is why card acceptance costs what it does.
- Scheme fees. What the network itself charges issuers and acquirers for coordination, brand, and settlement services, separate from interchange.
- Settlement finality. The legally recognised moment a payment becomes irrevocable; the property that matters more than speed, covered in the finality chapter.
- Chargeback liability. The allocation of who eats a disputed or fraudulent transaction; under card rules this mostly sits with issuers and merchants, not the consumer.
- On/off-ramp spread. The all-in cost of converting fiat to a stablecoin and back, including exchange fees and FX spread; usually the dominant cost of a stablecoin payment.
- Travel rule. The FATF requirement that originator and beneficiary information accompany transfers between virtual-asset service providers, which adds compliance overhead to stablecoin flows.
- Tokenised deposit. A bank liability represented on a shared ledger; the money stays on the issuing bank's balance sheet inside the regulatory perimeter, unlike a stablecoin. See the tokenised deposits chapter.