Chapter 15 · Cross-Border Payments and Payments Vision 2028
Part II built a machine that moves ₹29 lakh crore a month at home, near-free, in seconds. This final payments chapter takes it to the border — where the world still pays six per cent to move money — and then closes the Part the way the RBI closes its own chapters: with the Payments Vision 2028 roadmap, read critically against every gap these seven chapters found.
🎯 Learning Objectives
- Explain why cross-border payments remain slow, costly and opaque — the correspondent chain, nostro and vostro accounts, layered fees
- Describe India's remittance economy: the corridors, the volumes, and the cost target the world is missing
- State UPI's verified international footprint — the two integration models and the countries live as of September 2026
- Describe the rupee settlement routes — RDA, SRVA and RuPay abroad — and what they are for
- Close-read Payments Vision 2028: its lineage, theme, focus areas and specific commitments
- Evaluate the roadmap against Part II's findings — CILO 4's comparative judgement, applied to our own central bank
1. The Border as the Last Friction
Everything Part II solved was domestic. Cross the border and the machine reverts a century: transfers take days, cost several per cent, and no participant can tell the customer what she will finally receive. The standard global cost of sending a $200 remittance still runs around six per cent — twice the UN's Sustainable Development Goal target of three per cent — and the poor corridors cost most.
The villain is structural, not villainous. A cross-border payment moves through correspondent banking: your bank does not hold an account in the destination country, so it uses a correspondent that does — money actually moves between nostro and vostro accounts (accounts banks hold with each other in foreign currency), while SWIFT carries the instructions. Each hop adds a fee, a cut, a compliance check and a business day; each hop's participants know only their neighbours; and the chain reprices the currency at every step. Chapter 9's one-switch instant debit-and-credit has no international equivalent — yet.
For India this is not an abstraction. India is the world's largest recipient of remittances — inflows have run well above $100 billion a year in recent years — flowing chiefly from the Gulf's labour corridors (the UAE, Saudi Arabia) and the services corridors (the United States, the United Kingdom). Every percentage point of remittance cost is a tax measured in billions on some of the lowest-paid money-movers in the world. That is the prize the rest of this chapter walks towards.
2. UPI Goes Abroad
The international arm is NIPL — NPCI International Payments Limited, incorporated in April 2020, and it exports two different products that get conflated in headlines.
Model one: acceptance for Indian travellers (P2M). An Indian pays a foreign merchant by UPI QR from her Indian bank account in rupees — the foreign acquirer handles the local side. Live since Bhutan (July 2021), then Singapore's merchants (August 2021), the UAE (April 2022), and a cluster in February 2024 — France (Lyra; the Eiffel Tower was famously the first merchant, with roughly 9,000 transactions totalling about €9,30,000 by mid-2026), Mauritius and Sri Lanka. Nepal followed in March 2024, Qatar in late 2025, and Cambodia and Uzbekistan in 2026.
Model two: fast-payment-system linkages (P2P) — the deeper integration, connecting UPI to another country's real-time rail so account-to-account transfers cross the border instantly on mobile numbers or VPAs. The pioneer is Singapore's PayNow, live since 21 February 2023; Nepal's NCHL and Greece (via Eurobank) added P2P in 2026, and the Maldives' Favara linkage went live on 30 July 2026. By September 2026 UPI was reachable across roughly thirteen country-market arrangements, and the IMF's June 2025 assessment credited UPI with about 49 per cent of global real-time payment volume — the domestic machine, measured worldwide.
The direction of travel is diplomacy as much as finance: model-one deals follow heads-of-state visits, and 23 countries have signed India Stack / DPI cooperation MoUs. This is Chapter 5's export question answered in installments — and Chapter 29's global map will judge whether UPI-abroad is infrastructure or influence.
Sources: NIPL and NPCI releases; PIB factsheets and Lok Sabha replies (July 2026); press reporting to September 2026. Acceptance is at participating merchants only — "live in a country" never means every shop.
3. The Rupee Routes
Around UPI's travels sits an older, quieter project: settling more of India's cross-border business in rupees. Three instruments carry it. The Rupee Drawing Arrangement (RDA) — the classic remittance channel, where exchange houses abroad hold rupee accounts with Indian banks and remittances land without correspondent hops. The Special Rupee Vostro Account (SRVA) framework, created in 2022, which lets partner countries' banks open rupee accounts with Indian banks to settle trade — imports paid in rupees credited to the exporter's vostro, export proceeds routed back through it — a circuit that works without dollars in the middle. And RuPay abroad: fully interoperable with Bhutan in two phases (2019-2020), accepted through partners elsewhere.
State the motive precisely, because the seminar version oversells it. These routes cut cost and settlement time for users; they also give rupee-trade settlement an infrastructure in a sanctions-heavy world, and they matter most for the bilateral trade that dollar correspondent chains serve badly. They have not de-dollarised anything at scale — the rupee's share of global payments remains small. The honest framing: the rupee routes are optionality, built cheaply, before the day they are needed. Whether that day comes is a Chapter 29 argument.
4. Payments Vision 2028: A Close Read
The RBI has published payments visions since 2001; the lineage matters. Vision 2018 (2016-18) coincided with digital take-off and UPI's launch; Vision 2019-21 promised an "empowering exceptional e-payment experience"; Vision 2025 (2022) pushed inclusion and innovation deeper. Each document declared the last decade's frontier solved and named the next one. The Payments Vision 2028, issued 27 March 2026 with a horizon of December 2028, does the same — and its diagnosis is the most honest sentence the series has produced: the challenge "is no longer one of expansion of reach alone, but of deepening trust in payment systems, reinforcing resilience and expanding their global footprint."
The theme — "Shaping India's Payment Frontier" — organises four focus areas: user empowerment, safeguards against fraud, efficiency of cross-border payment frameworks, and ease of doing business, with an AI-led, data-driven approach across the ecosystem and risk-based regulation guided by innovation and inclusion. The concrete commitments this book cares about: a shared-responsibility framework in which the customer's bank and the beneficiary's bank jointly bear fraud liability — a direct answer to Chapter 14's mule-race asymmetry; issuer-channel toggles to enable or disable each payment mode; a Cyber Key Risk Indicators framework for non-bank payment operators; and on the border, a commitment to a cross-border framework that is "efficient, transparent, resilient and accessible", with work on the beneficiary leg's frictions and published progress reports.
Source: Payments Vision 2028, RBI, 27 March 2026 (rbi.org.in); RBI Annual Report coverage. Read the document itself in the hands-on — this section is the map, not the territory.
5. Reading the Roadmap Against the Gaps
Now the session's CILO 4 skill, turned homeward: evaluate a roadmap the way you would evaluate any jurisdiction's. The method — map each commitment to a documented gap; credit what fits; name what is missing — applied to our own seven chapters.
| Part II's gap | Vision 2028's answer | Verdict |
|---|---|---|
| Fraud and the mule race (Ch 14) | Shared issuer-beneficiary liability; mode toggles; cyber KRIs | Direct fit — the strongest commitment in the document |
| The border's six-per-cent cost (this chapter) | Cross-border framework; beneficiary-leg work; published reports | Right direction; no cost target stated |
| Cost recovery on a free rail (Ch 12) | Not a Vision item — answered by the October 2026 MDR settlement instead | Handled outside the roadmap; watch the November review |
| App-layer concentration (Ch 11) | Silent — the 30 per cent cap question is unresolved as its deadline arrives | The document's largest omission |
| Settlement as working capital (Ch 13) | Ease-of-doing-business stream touches it | Implicit at best |
| Data literacy in the ecosystem (Ch 10) | AI-led, data-driven approach; analytics as strategic assets | Aligned, though aimed at the RBI more than at users |
The pattern a close reader should notice: the Vision is strongest exactly where Chapter 14's victims are — trust and fraud — and quietest exactly where the political economy is hardest — competition. That is not a criticism unique to our central bank; roadmaps everywhere overweight technical fixes and underweight structural ones, because technical fixes antagonise nobody. Your evaluation earns its marks by saying so, with the table as evidence.
6. Part II, Closed
Seven chapters, one machine. Chapter 9 opened the architecture — PSPs, VPAs, the traced payment, on-us against off-us. Chapter 10 taught the data discipline the machine demands. Chapter 11 found the free rail concentrating and the cap contesting it. Chapter 12 priced the rail — interchange, RuPay, and the October 2026 return of MDR. Chapter 13 crossed the counter into acquiring's thin margins. Chapter 14 walked the dark side — fraud, mules, and the liability ladder. This chapter took the machine to the border and read the regulator's own scorecard for it.
The through-line of the whole Part is the one Chapter 8 stated: payments became a near-free public utility, and value migrated up the stack. So the book now follows the value. Part III is credit — the deepest revenue pool in finance — beginning where every loan begins: the value chain from origination to recovery, and the licences that govern each link. Chapter 16 opens Part III the way Chapter 9 opened this one: by tracing one journey end to end.
🛠 Hands-On: The Policy Document Close Read
This is the session's classroom activity, made self-paced. You will read Payments Vision 2028 itself, map it to Part II, run the corridor arithmetic, and stress-test the shared-liability proposal.
Steps
- Read the document. Download Payments Vision 2028 from the RBI website (Vision Documents page, 27 March 2026). Find three concrete commitments not listed in Section 4, and write one sentence on each.
- Map it. For each of your three commitments, name the Part II chapter whose gap it addresses — or state that it addresses none, and why that might be.
- Corridor arithmetic. A Dubai worker sends ₹1,00,000 home. Legacy wire costs about 6 per cent and takes two days; a linked-rail transfer costs about 1.5 per cent and arrives in minutes. Compute the saving, and the annual saving if she remits monthly.
- Stress-test shared liability. Under the Vision's proposal, the customer's bank and the beneficiary's bank jointly bear fraud liability. In 150 words: who clearly wins, who bears new cost, and what behaviour might a bank change at the margin to shift losses to the other side?
- The brief. In 200 words: does Vision 2028 answer the payments system's real gaps? Name the two biggest omissions (the scorecard table is a starting point, not an answer key), and state what you would add to the document if you could append one page.
View Solution / Walkthrough
1-2. Strong answers pick concrete, verifiable items (a named timeline, a framework, a pilot) rather than adjectives, and map them honestly — including to "no chapter", which is itself a finding: a Vision item with no gap behind it is either foresight or padding, and saying which is the exercise.
3. Legacy: 6% × ₹1,00,000 = ₹6,000 per transfer. Linked rail: 1.5% = ₹1,500. Saving: ₹4,500 per transfer; ₹54,000 a year at monthly frequency — about half a month's Gulf-side wage for many workers, from one switch of rail. Add the two days of working capital (Chapter 13's lens) and the case compounds.
4. Winners: customers (a deep-pocketed pair of banks replaces a freeze race) and honest beneficiary banks (mule accounts leave their books as onboarding tightens). New cost bearers: beneficiary banks, which now own losses their depositors' accounts received. Gaming risks: beneficiary banks delaying credits to push liability timing; issuers arguing every APP case is the beneficiary's mule; both sides litigating "joint" into 50-50 arithmetic the customer never sees. The strongest answers note the precondition — a shared mule-detection data layer, otherwise joint liability just splits the same losses.
5. Model briefs credit the fraud and cross-border commitments, then name the omissions: competition and concentration (the cap unresolved at its own deadline), and cost recovery's governance (the MDR question settled outside the Vision, by committee, twice deferred). The one-page appendix most worth adding: dated, numeric targets — remittance cost by corridor, fraud-loss rates, complaint-resolution times — because a vision without numbers cannot be audited. That is Chapter 10's lesson, aimed at the regulator.
Key Takeaways
The border reverts the machine a century: correspondent chains, nostro/vostro hops and layered fees — six per cent against a three per cent target, on the world's largest remittance inflow.
UPI abroad runs on two models — traveller QR acceptance (Bhutan 2021 onwards) and fast-payment linkages (PayNow 2023, Nepal, Greece, Maldives) — across ~13 arrangements by September 2026.
The rupee routes — RDA, SRVA, RuPay abroad — buy settlement optionality cheaply; they have not de-dollarised anything at scale, and honesty requires saying so.
Vision 2028 (27 March 2026, "Shaping India's Payment Frontier") shifts the frontier from reach to trust, resilience and global footprint — with shared bank liability for fraud as its sharpest commitment.
Close-reading the roadmap: strongest where fraud's victims are, silent where the political economy is hardest — competition. A vision without numbers cannot be audited.
Test Your Understanding
Choose one best answer per question. A calculator is allowed for the corridor arithmetic; the verified footprint of Section 2 is the data. Aim for 24/30 before entering Part III.
1. Under correspondent banking, money actually moves between:
2. In the cross-border chain, SWIFT's role is to:
3. The global average cost of sending a $200 remittance is about: (Calculator allowed.)
4. India's position in global remittances is:
5. India's remittances flow chiefly from:
6. UPI's international arm, NIPL, was incorporated in:
7. The first country to go live with UPI QR acceptance for travellers was:
8. The deeper model — linking two countries' fast-payment systems for instant P2P transfers — first went live with:
9. France's first UPI-accepting merchant, live from January 2024, was:
10. As of September 2026, UPI was reachable across about how many country-market arrangements?
11. The IMF's June 2025 assessment credited UPI with about what share of global real-time payment volume?
12. The Special Rupee Vostro Account framework exists to:
13. The honest verdict on the rupee settlement routes is that they:
14. Full two-way RuPay interoperability — each country's cards usable in the other — was achieved with:
15. Payments Vision 2028 was issued on:
16. Payments Vision 2028's theme is:
17. Vision 2028 states the ecosystem's challenge is no longer reach alone, but:
18. Vision 2028's shared-responsibility framework proposes that fraud losses be jointly borne by:
19. Vision 2028's stated method for the ecosystem is:
20. Among the measures introduced alongside the Vision period are:
21. The number of countries with India Stack / DPI cooperation MoUs is about:
22. On a ₹1,00,000 remittance: 6 per cent legacy wire versus 1.5 per cent linked rail saves: (Calculator allowed.)
23. At monthly frequency, that corridor saving compounds to: (Calculator allowed.)
24. The close-read finds Vision 2028 quietest about:
25. The general pattern this roadmap shares with vision documents everywhere:
26. The appendix most worth adding to Vision 2028, per the chapter's argument:
27. The traveller-QR acceptance model abroad primarily serves:
28. The caution to attach to every "UPI is live in country X" headline:
29. The through-line that carries the book from Part II to Part III:
30. Part III opens, in Chapter 16, by tracing: