FintechSphere · Research · Real time payments

Your money now moves in seconds. Who pays for that, and who is in control?

The best payment is the one you never notice, and that is exactly why it is hard to judge. Whoever designs these rails decides who earns, who carries the risk, and which countries keep control of their own money.

For banks and operatorsWhere the margin moves when the transfer is nearly free.

For merchantsWhy the bill is not the network fee, and what the rest buys.

For countriesWhat to keep at home, and what to share across borders.

Start reading

How to read this page. Nine chapters, with the shapes moving as you scroll. Pick how deep you want to go. The switch follows you down the page, bottom left.

Scroll

Chapter 1 of 9 · One payment

You scan. Both screens say success.

CustomerRM100.00Payment successful

MerchantRM100.00Payment received

A customer scans a merchant's QR code. The phone buzzes. The till chimes. It took about as long as a breath.

Now ask what a senior payments adviser asks. Have the banks settled? Who paid for the service? Whose rules made this safe and legal? We follow this one illustrative RM100 payment through the story and come back to it at the end.

The takeawayA success screen tells you the payment was accepted. It does not tell you the work is done.

Chapter 2 of 9 · Instant is not one event

Fast describes one moment, not the whole payment.

You see one number: your balance. The banks keep another: what they owe each other. Your screen can show success in seconds while the banks settle later. Both are true, on separate ledgers.

What the customer sees

What the banks settle

The takeawayAlways ask what became instant. The answer decides who carries the risk in between.

Deep diveThree consequences of separating what you see from what banks settle

Deep dive · Chapter 2

Three consequences of separating what you see from what the banks settle.

Speed is a property of specific events. Once you accept that, three practical questions follow, and each one has a different owner inside a payment system.

Risk

Someone carries the gap.

If a bank credits its customer before the banks have settled between themselves, someone is exposed for the gap. Who it is, and how it is controlled through prefunding, limits or timing, depends on the design. Ask which design a rail uses before asking how fast it is.

Operations

Unknown is a real status.

When a message goes out and no answer comes back, the payment may still have succeeded. Treat a timeout as a failure and the customer pays twice. Mature rails answer with a status enquiry, a stored outcome and a retry, and send what is left to manual review.

Open questions. How long should a system wait before it asks? Who decides when a retry is safe? What does the customer see in the meantime?

Protection

Final is not the same as protected.

Settlement finality says the banks cannot unwind the transfer. It does not say whether a customer who was tricked gets money back. Finality and protection are two separate promises, written in different rulebooks.

Three designs side by side

Design	What the customer gets	How the banks settle	The question to ask
UK Faster Payments	Money available around the clock	Net obligations settle three times each business day in Bank of England RTGS	How is the exposure between settlements controlled?
Australia NPP	Money available around the clock	Each payment settles individually through the RBA's Fast Settlement Service	What does always on settlement demand of every bank's treasury?
US RTP	Immediate availability	Settles against positions backed by a prefunded joint account at the New York Fed	How much liquidity must sit idle, and who funds it? Public guide dates from November 2020, so check current rules.

Back to the story: Chapter 3 ↓

Chapter 3 of 9 · Daily life

Payments go invisible when nothing in your day waits for them.

Invisible does not mean unprotected. Speed brings new jobs: a wrong recipient, an authorised scam, a timeout where nobody knows if it worked.

The takeawayA rail is judged on its worst day, not its best.

Deep diveSeven ways a fast payment goes wrong

Deep dive · Chapter 3

Seven ways a fast payment goes wrong, and what the rail owes you.

Pick a failure. Each one needs a specific capability that must exist before the first payment is made. This is the operating discipline that lets a payment go invisible.

What the rail must provide

Who is usually responsible

Back to the story: Chapter 4 ↓

Chapter 4 of 9 · The numbers

Big numbers, different rulers.

Three headline figures set the scale. Each comes from a different report with its own definitions, so read them as context, not as a scoreboard.

$2.6TGlobal payments revenue in 2025, up 3%, about 41% of banking revenue (McKinsey Global Payments Report 2026).

137Countries with 24/7 instant payment services by June 2026 (BCG).

56%Digital wallet share of global e-commerce value in 2025, against 33% in store (Worldpay Global Payments Report 2026).

With analysis on, the deep dive below breaks the numbers down by context: online, in store, transfers, Southeast Asia, growth and institutions.

The takeawayMore payments is not proof of a better system. Compare like with like, or do not compare.

Deep diveWhere payments happen, and how fast the pool is growing

Deep dive · Chapter 4

Where payments happen, and how fast the pool is growing.

Browse the key statistics by payment context. Every card shows its year, what it measures and where it came from. Global reports differ in definitions, so figures from different reports are never added together.

The takeawayChoose the context before you quote a number. A wallet share online and a rail share in store answer different questions.

Online versus in store: digital wallet share of payment value, 2025

Source: Worldpay Global Payments Report 2026, 42 markets. The report counts card funded wallet payments as wallets, so some wallet growth is card spending moving into apps. India figures are for UPI driven wallets.

The takeawayIn most markets the screen people touch is a wallet, so the fight for the customer happens in the app, not on the rail.

Southeast Asia, 2025: how people actually pay

Digital walletsDirect account to accountCards, directCash and other

Source: Worldpay Global Payments Report 2026, share of transaction value. Cards combine credit, debit and prepaid. Cash and other is cash and cash on delivery, plus BNPL online. Wallets include card funded wallets, and account payments made inside a wallet count as wallets, so the account to account bar understates the rail. Rows may not sum to 100 because of rounding.

The takeawayThailand and Malaysia pay by account far more than their neighbours, and Singapore stays with cards. One region, three different rails, so no single playbook fits.

Cash retreat: share of in store spend paid in cash, 2019 and 2025

Source: Worldpay Global Payments Report 2026. Asia Pacific average in 2025 was 13%. The report links the fall to national rails and QR codes: PromptPay in Thailand, BI-FAST and QRIS in Indonesia, DuitNow and the Financial Sector Blueprint in Malaysia, VietQR in Vietnam, UPI in India. It does not isolate how much each contributed.

The takeawayCash fell fastest where a national rail and QR arrived together. In these markets the real competitor is cash, not cards.

Growth: the payments revenue pool

McKinsey and BCG show revenue pools in trillions of US dollars, with different definitions, so compare each with itself only. Worldpay shows payment value, which is what consumers spend, not what the industry earns. The two right hand groups share one scale. 2030 bars are forecasts.

The takeawayVolume grows faster than revenue, so margin moves away from carrying the payment and towards the services around it.

What the data says, and what it does not

Reading 1

Growth is slowing while volume keeps rising.

McKinsey puts 2025 revenue growth at 3% against 9% a year over 2020 to 2025, and forecasts about 4% a year to 2030. BCG expects Asia Pacific revenue growth to trail volume growth as lower monetising rails gain share. Instant rails bring volume without bringing the same revenue per payment.

Reading 2

Wallets lead the interface, rails sit underneath.

Wallet share is high in Asia Pacific, at 77% online and 63% in store. That describes the app people touch, not the rail that moves the money. Brazil's Pix, at 42% online and 34% in store, and Thailand's PromptPay, at 44% and 43%, show account based rails competing directly for checkout.

Reading 3

Account to account is gaining where it is cheap and universal.

McKinsey's exhibit outlook has account to account payments overtaking cash in every region by 2035, and notes UPI and Pix absorbing cash at minimal fees. Cards hold their ground in North America, the UK and Western Europe.

Reading 4

In Southeast Asia, cash is the competitor.

In store cash fell from 64% to 22% in Malaysia, 68% to 30% in Thailand, 77% to 36% in Indonesia and 85% to 33% in Vietnam between 2019 and 2025. Cards did not do this work in these markets. National rails and QR codes did, which is why the debate here is about wallets and account payments, not card share.

Reading 5

Definitions hide the rail.

Worldpay counts a card funded wallet payment as a wallet payment, and UPI inside a wallet as a wallet payment. Direct account to account is 7% online and 4% in store globally, yet India shows 68% wallets with UPI underneath. Wallet share measures the interface. To measure the rail you need the operator's own data. Worldpay is also an acquirer, so its categories follow how a merchant accepts payment.

Reading 6

Fast on the wire, slow to the person.

Swift reports 75% of payments reach the beneficiary bank within ten minutes. The FSB data summarised by The Paypers has 35.4% of retail cross border payments reaching the recipient within an hour, against a 75% target, and an average cost of about 2.6% against a 1% target. These are different measures, and a likely reason for the gap is the last mile, where domestic instant rails matter.

The takeaway

The money is moving to the app, the rail underneath is getting cheaper, and the value is moving to the services around the payment.

Three things hold across every report we read. Wallets own the screen. Account rails take share where they are cheap and universal. Revenue grows slower than volume, so the margin sits in fraud, disputes, data and reconciliation more than in the transfer itself.

Questions to ponder

If revenue grows slower than volume, who pays for the resilience of the rail?
When wallets sit on top of account rails, who owns the customer and who owns the data?
Which markets could publish a like for like ruler, and who would be willing to?
Is cash retreating because the rails are good, or because incentives were paid to move people?

Back to the story: Chapter 5 ↓

Chapter 5 of 9 · How payments go invisible

You touch one layer. Five more do the work.

The top layer is not one thing. A scan, a tap, a card, a face or a phone number all start differently, then converge on the same chain. Pick a layer to light it up.

The takeawayInvisible does not mean simple. It means the complexity has been moved somewhere you do not have to look.

Deep diveFrom the first touch down to settlement

Deep dive · Chapter 5

From the first touch down to settlement.

Choose how the payment starts. The chain below shows which rail it usually rides and what each layer does. Two corrections matter: a card is an instrument, not an interface, and face or fingerprint is a way to prove it is you, not a payment rail.

Card rail

Merchant's acquirerReceives the card or token from the merchant and asks for approval.

Card networkRoutes the authorisation request and applies its own rules.

IssuerYour card bank. Checks funds and risk, then approves or declines.

Clearing and settlementDone through the card network's own arrangements, later than the approval.

Account rail

Merchant's bank or providerValidates the QR data and credits the merchant. Schemes use different names, such as acquirer or receiving participant.

Switch and schemeA national switch validates, routes and records settlement positions. Examples: Malaysia's RPP, Thailand's ITMX, India's NPCI, Indonesia's BI-FAST, UK Faster Payments.

Payer's bank or walletChecks the account and sends the instruction. Some schemes call it the issuer.

Interbank settlementDesign specific. Gross, prefunded or deferred net, as in chapter 2.

Every rail still needsRules and liability, fraud controls, dispute and refund handling, reconciliation, and a way to tell the customer the truth about status.

Who owns what

Different organisations at each step.

A processor, a merchant acquirer, an account holding bank, a direct participant and a settlement sponsor can all be separate firms. Terms like issuer and acquirer belong to a scheme's own rulebook. Use the scheme's words, not a card model for every rail.

On us payments

Sometimes the rail is never touched.

If payer and merchant bank with the same institution, the payment can stay inside it. Some schemes document this. In Malaysia's DuitNow QR flow, on us requests are not routed to the national switch, and other networks have similar on us rules. Rail statistics therefore understate total activity.

Where value sits

Between the layers.

The directory that resolves a phone number, the fraud signal, the reconciliation file and the dispute desk are the services merchants and banks actually feel. That is our first hypothesis for the end of this story.

Back to the story: Chapter 6 ↓

Chapter 6 of 9 · Who pays

If the network fee is tiny, why is the merchant's bill not?

The merchant never pays the network directly. The merchant pays its acquirer a merchant discount rate, the MDR. The acquirer pays the network and switching fee to the rail and, on most card payments, an interchange fee to the payer's bank. It buys fraud, dispute and operations services, adds its own cost, then adds a markup.

Merchantpays MDR toAcquirerpaysNetwork and switchIssuer, on cardsValue added services

MDR is not the network fee. It is the whole stack plus a margin the market sets. A low network fee does not mean low total cost, and the services around it are what keep the rail resilient and trusted.

The takeawayJudge a payment method on total cost of acceptance, not on one line of the invoice.

Deep diveBuild the merchant's bill, one layer at a time

Deep dive · Chapter 6

Build the merchant's bill, one layer at a time.

Every figure here is synthetic and chosen to show behaviour, not to quote any acquirer or scheme. Switch between an account to account QR payment, which usually has little or no interchange, and a card style payment, which does. Then turn services on and off and watch what happens to the MDR and to the losses nobody invoices.

Network and switching feeOne fee, paid to the rails that carry and process the message, the network and the switch.

Interchange feePaid to the issuer, the payer's bank, for funding the payment and carrying the risk. Account rails often have none.

Payment

Ticket size

Paid to the rail

Network and switching fee, share of value

Paid to the issuer

Interchange fee, share of value

Value added services the acquirer provides

Share of payments disputed

Fraud attempts, share of value

The acquirer

Operating cost per payment

Markup, set by competition

Merchant pays (MDR)

Paid to the rail

Safeguards on

Where each RM of the bill goes

Synthetic model. The network fee covers the network and the switch. Services cost per payment: fraud screening 1.0 sen, refund management 0.8 sen, reconciliation 0.5 sen, round the clock support and monitoring 1.2 sen. Dispute handling costs RM8 per dispute. With screening on, 30% of fraud attempts get through. With dispute handling on, 20% of disputed value is lost, against 100% without it.

Message 1

Low network fee, not low total cost.

Push the network fee to zero and watch the share it represents. For small tickets the fixed pieces dominate; for large tickets the percentage pieces do. The network is rarely the biggest line.

Message 2

Services buy resilience and trust.

Fraud screening, dispute handling, refunds, reconciliation and round the clock support are what a merchant relies on when something goes wrong. Switch them off and the invoice shrinks while the exposure grows.

Message 3

The market sets the markup.

Schemes and regulators can shape the network fee. The acquirer's markup is competitive. Regulators sometimes steer fees directly. The Reserve Bank of Australia regulates card interchange, and Bank Negara Malaysia said the DuitNow QR merchant fee should stay at or below debit card fees, with banks waiving it for micro and small businesses.

Back to the story: Chapter 7 ↓

Chapter 7 of 9 · The border

At the border, instant meets two currencies.

Two different products cross borders, and each can be linked two ways. Keep them apart or the conversation turns to mush.

TransfersPerson to person. Bilateral pairs such as PayNow with PromptPay or UPI. Multilateral: Project Nexus.

QR paymentsAt the merchant. Bilateral: QRIS, PromptPay, KHQR, DuitNow QR and UPI link pair by pair. Multilateral: commercial networks such as Alipay+.

The catchTwo kinds of multilateral: Nexus is owned by central banks, Alipay+ is a commercial network.

The takeawayNever call a pilot a network. Label every corridor live, announced or being built, and date it.

Deep diveThe map of who is connected to whom

Deep dive · Chapter 7

The map of who is connected to whom.

Read this firstEverything on this map is compiled from information found online: central bank and operator press releases, news reports, and secondary sources such as Wikipedia and regional trackers. We have not verified any link with the operators. Dates conflict between sources, some links may have changed since, and some are only reported once. Treat the map as a starting point for questions, not as a record. Verify every corridor with the central banks or scheme operators before relying on it.

Pick a country to see the network from its point of view. Every line is a link between two national systems, or between a national system and a multilateral hub. Lines are dashed when a link is announced, only under discussion, or reported by a single secondary source.

Two products, two models

Bilateral: one link per pair

Multilateral: connect once, reach many

Transfers
person to person

PayNow with PromptPay, UPI and DuitNow

Singapore's PayNow links to Thailand's PromptPay (April 2021), India's UPI (February 2023) and Malaysia's DuitNow. Users enter a mobile number or proxy in their existing bank app, and the sender sees the FX rate upfront.

Each pair is its own project with its own participants, FX arrangement and rules.

Project Nexus

Nexus standardises how instant payment systems connect. Each system connects once and reaches every other country in the network, with a target of arrival within 60 seconds in most cases.

Nexus Global Payments, a non profit in Singapore, was founded in 2025 by the central banks of India, Malaysia, the Philippines, Singapore and Thailand. Bank Indonesia joined in February 2026. We found no confirmed date for the first live payment.

QR payments
at the merchant

QRIS, PromptPay QR, KHQR, VietQR, DuitNow QR, SGQR+ and UPI

National QR schemes link pair by pair. In the sources we read, Indonesia lists live QRIS links with Thailand, Malaysia, Singapore, Japan, South Korea and China. The Bank of Thailand lists live QR links with eight markets. Cambodia reports five live KHQR links in ASEAN: Thailand, Laos, Vietnam, Malaysia and Singapore. India's UPI is accepted in a growing list of countries from the Gulf to Southeast Asia.

A traveller scans the home app at the local merchant's code.

Alipay+ (Ant International) and similar networks

One commercial network connects partner wallets and banks from many markets to the merchants that accept it. Worldpay describes Alipay+ as serving users in 14 markets, and notes Tenpay Global as a similar platform from WeChat Pay's parent.

Inbound is live in markets such as Malaysia. The August 2025 agreement with PayNet says Malaysians will be able to pay at over 100 million Alipay+ merchants. Outbound is announced, not live.

Six things the map tells a payments leader

1 · Two kinds of multilateral

Public network or commercial hub.

Nexus is owned by central banks, non profit, and built to connect domestic instant payment systems. Alipay+ is a commercial network operated by Ant International. They differ in who sets prices, who sees the data and who can change the rules. That is a sovereignty question, taken up in chapter 8.

2 · Different jobs today

One is live, one is being built.

Alipay+ delivers inbound tourist spending at scale now. Nexus is aimed at account to account transfers between systems and is still being built, with no confirmed first live date in the sources we found. Press reports differ between 2026 and 2027.

3 · A mesh built pair by pair

4 · Cross border is still small

Where we can measure it, it is under 1%.

One operator reports 29.7 million cross border QR transactions in Malaysia in 2025, against 8.44 billion digital transactions, roughly 0.35%, and 2.5 times the year before. We found no comparable figures for most other corridors, which is itself a finding. The perimeters differ, so treat it as scale, not share.

5 · Direction matters

Inbound and outbound are different projects.

Letting visitors pay at home is easier to scale than letting residents pay abroad, because it needs one set of merchants to accept one more app. Outbound needs your banks and wallets to connect, price FX and handle disputes in a foreign market. Korea reaches Malaysia through Alipay+, not through a published scheme to scheme link.

6 · What stays hard

The middle of the journey.

A shared QR format is only a start. Each link still needs agreement on messaging, currency conversion, sanctions and money laundering screening, settlement sequencing, beneficiary verification and who handles a dispute. Nexus gateways are designed to take on messaging and screening centrally.

Why connect once beats connect to each

Payment systems to connect

Separate bilateral links, worst case

Connections to a hub

Fully meshed bilateral links grow with the square of the number of systems: n times n minus 1, over 2. A hub grows in step with n. Real networks do not build every pair, so this is the ceiling, not a forecast. But it shows why central banks chose to build Nexus rather than fifteen separate projects, and why a hub also concentrates power in whoever runs it.

Corridor register

Showing the country selected above, or everything if none is selected.

Corridor	Product	Schemes	Model	Date	Status	Source

Status key. Live: reported as operating in at least one source. Announced: agreement or plan without a live date. Exploring: talks only. Reported: one secondary source only, not confirmed elsewhere. Being built: under development. This register is compiled from online sources and has not been verified with the operators.

Back to the story: Chapter 8 ↓

Chapter 8 of 9 · Standards and sovereignty

Who writes the rules everyone else follows?

Cards had a head start. EMVCo, owned by six card networks, publishes the technical specifications for chip, contactless tap, tokenisation and online authentication. That is why one card or phone works almost anywhere.

Account rails grew up at home. Domestic schemes were built for domestic needs. Now that interoperability is the goal, account to account payments need their own shared rule set. And every country has to decide which parts of its payment system it must control itself.

The takeawayDecide which rules must be identical everywhere, and which can stay local. Everything else is detail.

Deep diveThe non negotiables for account to account interoperability

Deep dive · Chapter 8

The non negotiables for account to account interoperability.

The card world solved a technical coordination problem long ago. It did not solve everything, and account rails need their own answer. Below is what EMVCo does and does not do, then our working list of what has to be standard.

What EMVCo is, and is not

Does

Maintains technical specifications.

Owned by American Express, Discover, JCB, Mastercard, UnionPay and Visa. It maintains the EMV specifications, free to use, and approves products against them. They cover contact and contactless chip, tokenisation, 3-D Secure, secure remote commerce and a QR code specification.

Does not

Set commercial rules.

EMVCo does not mandate that anyone uses its specifications, and it does not set implementation requirements. Payment networks decide those for issuers, merchants and acquirers. Prices, liability and dispute rules stay with each network.

The takeaway

Streamlined because specs are shared, not because rules are.

The experience feels uniform because the technical handshake is shared. Each network still writes its own commercial and liability rules. In short, EMVCo standardises the plumbing, not the whole rulebook.

The QR layer is partly standard already

Exists

EMV QR code specification and mark.

EMVCo's QR specification covers merchant presented and consumer presented modes and standardises the data in the code. What happens after the scan is left to the payment provider. EMVCo also offers a royalty free EMV QR Payment Mark and a QR Scan Icon so a consumer can recognise that a code is accepted.

Reality

Adoption varies, and it stops at the code.

Singapore's SGQR and Indonesia's QRIS build on the EMVCo foundation. Brazil's Pix uses an EMV based code. Thailand's QR standard is described as aligned. India's UPI QR is proprietary. Other national QR standards may or may not follow the EMV payload, so check each one. A shared payload still leaves messaging, FX, screening and settlement to be agreed link by link.

Our working list of non negotiables

What must be standard	Why it cannot be left to each pair	Where it stands
QR payload and acceptance mark	A customer must recognise a code and an app must read it anywhere.	Exists EMV QR and the EMV QR Payment Mark, adopted unevenly
Trust mark and merchant verification	An acceptance mark says a code is accepted, not that the merchant is genuine. Payers need verified merchant names on the confirmation screen.	Partial Several schemes show the merchant name before the payer confirms. A trust mark with one meaning across schemes is not established
Message language and minimum data	Each hop translates otherwise, and data gets lost or rejected.	Partial ISO 20022 is the common base, and the FSB lists harmonised requirements as completed. Nexus gateways translate between domestic formats, which shows the gap
Timeouts and unknown status	Without agreed waiting times and a mandatory status enquiry, one side retries while the other already paid. Unknown must never be treated as failed.	Proposed Domestic schemes define their own. We found no shared cross border rule. Nexus aims for arrival within 60 seconds in most cases
Alias resolution and beneficiary check	A payer should see who they are paying before they pay, in any corridor.	Partial Common in domestic rails, patchy across borders
Refunds, returns and disputes	Common reason codes and time limits, or a cross border dispute bounces between two rulebooks.	Proposed
Fraud and scam signals	Scammers choose the weakest corridor. Fast freeze and trace must work across schemes.	Partial Domestic recovery and information sharing exist in several markets, and some operators have agreed bilateral threat intelligence sharing. A shared protocol does not
Sanctions and money laundering screening	Who screens, on what data, and who is liable for a miss, decides how fast a corridor can run.	Partial Agreed link by link today. Nexus gateways are designed to take on screening centrally
Price and FX transparency	The all in price, including conversion margin, must be visible before the payer confirms.	Partial PayNow's links show FX rates upfront. In the FSB data summarised by The Paypers, 62.9% of services disclosed both cost and speed
Operating hours and settlement windows	Instant payments run around the clock across time zones, while many settlement systems do not.	Partial The FSB roadmap extended RTGS operating hours to widen overlap. Liquidity arrangements for nights and weekends stay local
Availability and incident reporting	A corridor is only as reliable as its weakest participant. Published uptime and outage reporting build trust.	Proposed Operators report availability for their own networks, with no common definition across schemes
Conformance testing and certification	A bank or wallet should pass shared test scripts before joining a link, so one weak participant does not break the corridor.	Proposed Card networks rely on certification labs. We found no equivalent shared practice for account rails
Finality and liability	When money is irrevocable, and who bears a loss when a corridor fails, must be written down in advance.	Proposed Nexus has a scheme and governance framework, published by BIS in July 2024

Status labels are our reading of public sources at the research cut off. The list is a proposal for debate, not an existing standard.

Three ways to get there

Option 1 · Standards body

An EMVCo for account rails.

A neutral technical body maintains shared specifications that any scheme can adopt, free of commercial rules. Fast to agree on the plumbing, silent on liability and price.

Option 2 · Hub rulebook

Nexus style scheme rules.

One governed scheme sets connection, FX, screening and liability rules for everyone who joins. Strongest consistency, and a new concentration of power for the operator.

Option 3 · Bilateral memoranda

Keep agreeing pair by pair.

Preserves local control and moves quickly for a single corridor. Does not scale, and every pair reinvents the timeouts and the dispute rules.

Next: sovereignty ↓

Deep diveSovereignty: six practical questions

Deep dive · Chapter 8, part two

Sovereignty: six practical questions.

Sovereignty is not a slogan. Pick a dimension, and the sphere lights the region it controls. For each one we give the strongest case on both sides, and where a sensible country lands.

Keep at home

Where it lands

Reference point: the Eurosystem's March 2026 payments strategy sets out an institutional view on resilience and autonomy. It is one policy position to weigh, not proof that one ownership model wins.

Own the control points you cannot afford to lose. Connect everywhere else.

Back to the story: Chapter 9 ↓

Chapter 9 of 9 · The next frontier

Back to the RM100.

The payment succeeded. The merchant now needs four things the rail does not give for free.

ReconciliationMatch every payment to a sale and handle the exceptions.

TrustKnow it is real, and know who to call when it is not.

LiquidityFunds that are usable, not only credited.

ReachCustomers and corridors beyond the home market.

The takeawayTest whether value is moving toward the services around the rail. Build, buy, connect or govern accordingly.