Module 2

Real-Time Payment Networks

How FedNow, Faster Payments, UPI, PIX, and SEPA Instant are reshaping settlement globally

---

The Global RTP Landscape: 70+ Live Networks in Operation

Real-time payment networks are not a future technology. They are operational today across six continents. Over 70 countries have live RTP networks processing significant transaction volume. The adoption curve is steep and accelerating. What was experimental five years ago is now structural infrastructure in the world's largest economies.

The geographic distribution matters because it reveals the regulatory and market forces driving adoption. Developed economies with mature banking infrastructure deployed RTPs first, but emerging markets are seeing faster adoption and higher volume concentration. This reflects different economic realities: developed markets have competitive pressures from existing card networks and need regulatory nudges. Emerging markets see RTPs as leapfrog infrastructure that bypasses the historical limitations of correspondent banking.

Real-time payment networks were engineered for the economics of 2020s commerce, not the batch settlement processes of the 1960s. Every major economy is deploying them. The question is not whether RTPs will dominate payments infrastructure. The question is how quickly the transition completes.

The networks vary significantly in architecture, coverage, and adoption model. Some are government-mandated. Some are market-driven. Some are built as utilities. Some are operated by consortiums of banks. Each approach creates different incentive structures and adoption curves. Understanding how each network works is essential to understanding how payments will flow in your market.

---

FedNow: The US Real-Time Payment System

Launch and Adoption Timeline

FedNow, operated by the US Federal Reserve, launched publicly on July 19, 2023. It was the first real-time payment system operated directly by a central bank in the United States. The announcement came with expectations of rapid adoption, but the reality has been more measured. As of early 2026, FedNow processes over 100 million transactions monthly and has reached an annual volume run rate exceeding $2 trillion.

The adoption curve has been steady but not explosive. US banks were initially cautious about regulatory requirements, integration complexity, and customer demand. Early adopters were primarily large banks and fintechs looking to differentiate on speed. But as the network effect accelerated through 2024 and 2025, adoption broadened. By early 2026, over 9,000 financial institutions have access to FedNow, representing approximately 90 percent of US bank deposits.

Volume growth is outpacing predictions. The Federal Reserve originally projected 10 to 15 percent annual growth. Volume has been growing 25 to 30 percent annually. This suggests that once access becomes universal and consumer awareness increases, adoption accelerates faster than expected. The network is still in early growth phase, with significant runway ahead as use cases expand beyond peer-to-peer transfers.

Technical Architecture and Capabilities

FedNow operates as a hub-and-spoke model with the Federal Reserve at the centre. Banks connect to FedNow infrastructure and send payment messages through the network. The system processes payments individually (not batched) and settles immediately. The payment is final and irrevocable once processed. There is no chargeback mechanism and no dispute resolution beyond what the receiving bank chooses to provide.

Each payment is capped at $500,000 per transaction. This limit was set by the Federal Reserve to manage systemic risk and encourage use cases in the consumer and small business space. Large commercial payments and wire transfers still route through Fedwire, the existing high-value payment system. The $500,000 limit is designed to prevent FedNow from competing with Fedwire, but as use cases for larger payments grow (particularly in business-to-business), there is likely pressure to raise the limit in the coming years.

The system is available 24 hours per day, seven days per week, 365 days per year. This is a fundamental shift from card networks (which have defined settlement windows) and most ACH systems (which process on business days). The round-the-clock availability means a payment on Saturday at 3 am arrives instantly on Monday morning, just as intended. This is the core value proposition for businesses that need to move money outside traditional banking hours.

Information provided with each payment is limited. FedNow transmits the basic elements: payer account, payee account, amount, and a reference field. It does not provide detailed remittance information or rich data about the payment purpose. For merchants, this means they must match payments to orders using the reference field, not detailed transaction data. This is a gap compared to ACH payments (which can include detailed transaction information) and something the market is working to solve through ancillary services.

Current Limitations and Challenges

FedNow has limitations that matter for specific use cases. First, there is no consumer protection framework. If a payment is sent to the wrong recipient by mistake, there is no automatic reversal process. The receiving bank must be contacted and must agree to reverse it. If the receiving bank is a different institution, there is no unified reversal protocol. This creates friction for accidental payments and puts burden on customers to resolve errors manually.

Second, fraud liability is unclear. If a customer's credentials are compromised and an attacker sends fraudulent payments through FedNow, who bears the loss? Card networks have clear chargeback processes. FedNow has no equivalent. Regulatory guidance is evolving, but the liability framework is not yet settled law. Banks are conservative about this uncertainty, which slows adoption for retail use cases where fraud protection matters.

Third, adoption is uneven geographically and by institution size. Large money centre banks adopted FedNow quickly. Community banks and credit unions have been slower. This creates fragmentation where some institutions offer FedNow and others do not. A customer cannot send a FedNow payment to an institution that has not connected to the system. Interoperability across all institutions is not guaranteed, which limits network effects.

Fourth, the $500,000 transaction limit constrains use cases. Large business-to-business payments, commercial real estate transactions, and institutional transfers cannot use FedNow. The market has to decide whether to process these through FedNow (requiring multiple transactions) or stick with Fedwire. Many institutions choose Fedwire to avoid operational complexity of splitting payments.

Where FedNow Wins

FedNow is ideal for use cases where speed and around-the-clock availability matter more than comprehensive consumer protection. Gig economy payments (where workers need access to earnings immediately) are a natural use case. Business-to-business invoicing where both parties have bank accounts is ideal. Treasury and working capital optimization where companies need to move money between accounts instantly. Bill payments, rent, and scheduled transfers are all strong use cases.

The real adoption acceleration will come when consumer applications integrate FedNow as a payment method. Imagine a delivery platform that offers instant earnings payout through FedNow. Workers see payment hit their account immediately, not next day. That user experience is powerful enough to drive adoption independent of other payment methods. Once that happens, network effects accelerate and FedNow becomes the default for certain payment types.

---

Faster Payments: The 18-Year Pioneer

The System That Proved Real-Time Works at Scale

The UK Faster Payments Service launched in 2008, making it the oldest real-time payment system in active operation. Eighteen years of operational data proves that real-time payments can work reliably at massive scale. The system processes around 20 million transactions daily and is approaching trillions in annual volume.

Faster Payments is notable because it predates and predicted the RTP wave. When it launched, the idea that payments could be instant seemed impossible. Batch settlement was accepted as law of nature. Faster Payments proved otherwise. The system was built by the UK banking industry to address the exact problems that are now driving RTP adoption globally: slow settlement, high costs, and friction in cross-border payments.

The success of Faster Payments directly influenced regulatory decisions in other countries. When the EU was designing SEPA Instant, they looked to Faster Payments as the operational model. When the US was designing FedNow, they studied Faster Payments infrastructure and learned from 18 years of operational lessons. Faster Payments is not the largest or the fastest-growing RTP network, but it is the proof of concept that made all others possible.

Architecture and Lessons Learned

Faster Payments operates as a three-tier network. Participating banks connect to the Faster Payments Service, which runs the hub. Member banks send payment instructions through standardised message formats. The system batches payments in short cycles (typically every 20 seconds) rather than processing one-by-one. This batching approach allows for more efficient processing than true real-time, while still delivering settlement within seconds rather than hours or days.

The 20-second batching window was an operational choice to balance processing load and settlement speed. True transaction-by-transaction settlement would require more computing resources and more complex reconciliation. The 20-second window provides near-real-time settlement without the complexity of true atomic processing.

Faster Payments caps transactions at £250,000 (approximately $315,000 USD), similar to FedNow's $500,000 limit. Like FedNow, the cap is designed to keep the system focused on retail and small business use cases. Larger payments route through CHAPS (Clearing House Automated Payment System), the UK's high-value payment system.

What Faster Payments Got Right and Wrong

The system got the basics right: instant settlement, 24/7 availability, low cost per transaction, and sufficient scale to drive network effects. After 18 years, the UK market has largely switched to Faster Payments for most domestic payments. Bank transfers that used to take three days now complete in seconds. This alone justifies the system's existence.

What Faster Payments got wrong is the limited information sharing. Like FedNow, the system provides minimal remittance information with each payment. This creates reconciliation friction for businesses that receive many payments and need to match them to invoices or orders. The Faster Payments Service has acknowledged this limitation and is working on newer formats that allow richer information, but the installed base still uses limited information payments.

The second challenge is fraud and consumer protection. Faster Payments has no chargeback mechanism. This has created a market for fraud where scammers convince people to send "refunds" or "transfers" through Faster Payments, knowing that once sent, the money is gone forever. The UK banking industry and regulators are working on a mandatory refund scheme, but it is not yet law. The absence of fraud protection has slowed retail adoption and created reputation damage for the system.

The third limitation is cross-border integration. Faster Payments is UK-only. When a UK payment needs to reach someone outside the UK, it still routes through correspondent banking and SWIFT. This means Faster Payments solves the domestic payment problem but not the international problem. This limitation motivated the EU to design SEPA Instant as a multinationally integrated system from the start.

---

UPI: The 10 Billion Transaction Per Month Model

Why India's Network Scaled Faster Than Others

India's Unified Payments Interface (UPI) processes over 10 billion transactions per month, making it the highest-volume real-time payment network globally. To put this in perspective, FedNow processes approximately 100 million transactions monthly (12 billion annually), while UPI processes more volume in one month than FedNow processes in a year.

UPI achieved this scale in under a decade. The system launched in 2016 and reached critical mass by 2020. The adoption curve was explosive and sustained. By 2026, UPI handles more payment volume than all Indian card networks combined. This is not a supplementary payment system. This is the primary payment infrastructure for Indian commerce.

The reasons for UPI's success are instructive. First, it solved a real and immediate problem. India had 1.2 billion people but only 50 million bank accounts and even fewer payment cards. UPI created a bridge between the banking system and consumers without traditional payment infrastructure. You could use UPI with just a basic smartphone and a bank account. The friction to entry was near-zero.

Second, UPI was designed for emerging market constraints. The network works on 2G connectivity. It handles frequent network interruptions. It does not require near-instantaneous communication. This architectural choice made UPI accessible to hundreds of millions of people in areas with unreliable internet. Most other RTP networks were designed assuming reliable high-speed connectivity, which excluded emerging market use cases.

Third, regulatory mandate and bank participation drove adoption. The NPCI (National Payments Corporation of India) designed UPI and mandated bank participation. All banks in India had to support UPI. There was no choice. This unified participation created instant network effects. Every payment goes through the same rails. There is no fragmentation into adoption tiers.

Technical Architecture and Interoperability

UPI operates as a virtual payment address system layered on top of the bank transfer infrastructure. Instead of sharing bank account numbers, Indians share a UPI address (typically name@bank or name@upi). When you send payment to a UPI address, the NPCI infrastructure routes the payment to the correct bank, settles it instantly, and confirms completion.

The brilliance of the UPI design is its simplicity and interoperability. The system is agnostic about which bank you use. A payment from one bank to another bank requires no correspondent banking, no SWIFT messages, no intermediaries. It is a direct bank transfer with instant settlement. The UPI layer handles routing and reconciliation transparently.

UPI also introduced QR code payments, which became central to the value proposition. A merchant displays a QR code. A customer opens their phone, scans it, confirms the amount, and sends the payment. The merchant receives instant confirmation. This user experience is so simple and frictionless that it drove mass adoption. Every transaction becomes as simple as a two-step process: scan, confirm.

The transaction limit is 100,000 Indian Rupees (approximately $1,200 USD) per transaction. This is higher than FedNow or Faster Payments in percentage terms relative to local purchasing power, reflecting the different use cases UPI was designed for. The network handles everything from street food purchases to large personal transfers.

Scaling Challenges and Solutions

UPI's success has created scaling challenges. Processing 10 billion transactions per month requires massive infrastructure investment. The NPCI has had to continuously upgrade servers, network capacity, and processing systems. Peak load periods (festivals, paydays) push the system to limits. During Diwali shopping season, UPI sometimes experiences brief outages or slow processing, which is a sign that the infrastructure is nearing capacity even with heavy investment.

The second challenge is fraud prevention at scale. With 10 billion transactions monthly across hundreds of millions of users and merchants, fraud patterns are harder to detect. The NPCI has implemented machine learning systems to catch fraud, but the volume means some fraud inevitably slips through. The customer protection framework is being strengthened, but it remains a work in progress.

The third challenge is standardisation. UPI enabled dozens of apps to build on top of the infrastructure. Google Pay, WhatsApp Pay, Amazon Pay, and others all use UPI for settlement. This innovation and choice is good for consumers, but it has also fragmented the UPI experience. Some apps provide better fraud protection. Some provide better customer service. The core UPI infrastructure is only as strong as its weakest participant app.

---

PIX: The 150 Million User Overnight Success

The Brazilian Instant Payment System That Broke Adoption Records

PIX, Brazil's instant payment system, reached 150 million registered users in three years, making it the fastest-adopted payment infrastructure in history. To put this in perspective, Brazil has 215 million people. PIX is used by two-thirds of the entire population. The adoption curve was so steep that it eclipsed Brazil's card networks in transaction volume by 2023 and has continued to grow at rates exceeding 50 percent annually.

PIX launched in November 2020, less than three months after the central bank announcement. The speed of deployment was unprecedented. The central bank mandated that all banks in Brazil provide PIX connectivity. The adoption was not optional or phased. It was full participation from day one. This eliminated all fragmentation and network effect delays. Every transaction was immediately available to reach any participant.

The economics of PIX drove rapid adoption. Card transactions cost merchants 2 to 3.5 percent in interchange and fees. PIX transactions are capped at 1 Real (approximately $0.20 USD) per transaction for merchants, or free for consumers. The cost difference is so extreme that merchants aggressively promoted PIX as a payment method. Over time, consumer behaviour shifted to match merchant incentives.

Architecture and Key Features

PIX is designed around the concept of the Instant Payment Account (Conta de Pagamentos Instantâneos). Every participant bank maintains a PIX account for each customer. Payments are routed directly through a central clearing house operated by Brazil's central bank. Settlement is instantaneous. The payment is final. There is no reversal mechanism (recently a mandatory refund window was added for fraud cases, but core transactions are irreversible).

Like UPI, PIX uses a QR code-based payment method as a primary interface. A merchant displays a static QR code. A customer scans it, confirms, and sends payment. The merchant receives immediate confirmation. The experience is so simple that it drove adoption even among elderly populations and people with limited smartphone skills. PIX was available to anyone with a phone.

PIX transaction limits are much higher than other RTP networks. There is a per-transaction limit of 5 million Reais (approximately $1 million USD) and a 20 million Reais per-day limit. These high limits reflect Brazil's decision to use PIX for all payment types, including large B2B transactions and institutional transfers. Brazil did not segment PIX into consumer and commercial versions. It built one system for all use cases.

Adoption Curve and Market Disruption

The adoption curve for PIX followed a pattern unseen before in payment infrastructure. In the first three months, PIX handled 10 million transactions. By month 12, it was processing 500 million transactions monthly. By month 36, over 150 million users were registered and the system was processing billions of transactions monthly.

This rapid adoption created market disruption that the card networks did not anticipate. Card transaction volume did not just grow slower than PIX. It declined. Consumers and merchants actively migrated off cards to PIX. Merchants that had been paying 3 percent in card fees were now paying near-zero. Consumers that had been waiting 2 to 3 days for transfers were now seeing instant settlement. The incentive to use cards disappeared for many use cases.

By 2024, PIX was processing more transaction volume than all of Brazil's card networks combined. This was not expected to happen until 2027 or 2028. The speed of adoption caught financial institutions and payment networks by surprise. The lesson is clear: when a government-backed instant payment system exists with lower costs and better user experience, market forces shift volume away from legacy infrastructure far faster than predicted.

---

SEPA Instant: The Multinational EU Integration

The European Approach to Real-Time Payments

The EU's approach to real-time payments is distinctive because it attempted to integrate across 27 countries simultaneously, rather than building country-by-country. SEPA Instant Credit Transfer (SCT Inst) was designed to enable instant payments between any two IBAN accounts in any EU country using a single standardised protocol.

This cross-border interoperability is the defining feature of SEPA Instant. Unlike PIX (which is Brazil-only) or FedNow (which is US-only), SEPA Instant is continent-wide. A payment from Germany to Portugal to Italy reaches instantly and settles in seconds. The same standardised clearing and settlement infrastructure applies across all 27 member states.

The challenge of achieving this was immense. Each country had existing national payment systems with different architectures, different rules, and different operator structures. The EU had to design a protocol and clearing infrastructure that all countries would adopt and operate identically. The result is SEPA Instant, which became mandatory for all EU banks on October 19, 2025.

Mandatory Participation and Adoption

SEPA Instant is not optional in the EU. The regulation is clear: every bank in the EU must offer instant payment capabilities by October 2025. Coverage must extend to 24/7/365 availability. The transaction limit is 100,000 Euros per transaction. The costs must be comparable to regular SEPA transfers (which are typically free or near-free).

The mandatory participation requirement means that by late 2025, every EU citizen and business has access to instant payments. There is no fragmentation by bank or country. This is the fastest path to continent-wide adoption any RTP network has taken. By early 2026, we are seeing the adoption curve accelerate as businesses and consumers discover that instant payments are now universally available.

The cross-border capability is transformative. Freelancers working across EU countries can now invoice clients and receive payment within seconds, rather than waiting 2 to 3 business days. Small businesses can optimize working capital by moving money between countries instantly. The cost of cross-border commerce drops materially when settlement is instant and costs are near-zero.

Coverage Gaps and Remaining Challenges

SEPA Instant solves intra-EU payments but not payments to or from outside the EU. A payment from the US to Germany still routes through correspondent banking and SWIFT, even though Germany-to-Germany payments are instant. This creates a two-tier system where domestic/intra-EU payments are fast and cheap, but anything crossing EU borders reverts to the old correspondent banking problems.

The second challenge is consumer fraud protection. Like other RTP networks, SEPA Instant has no built-in chargeback mechanism. A fraudulent payment cannot be reversed through the network. Regulatory guidance requires that fraud losses be covered by banks within certain limits, but the framework is still being clarified. This uncertainty has slowed retail adoption for some use cases.

The third challenge is information richness. SEPA Instant supports more detailed remittance information than FedNow or Faster Payments, but still less than what merchants would ideally like. A merchant receiving a payment needs to automatically match it to an invoice. SEPA Instant allows enough information for this in most cases, but additional standardisation is ongoing.

---

Technical Architecture: How Messages Move Across Networks

All modern RTP networks use similar technical architectures, though details vary. They all operate on standardised message formats (typically ISO 20022), they all have settlement finality within seconds, and they all provide end-to-end instant confirmation of payment completion.

The common flow is: payer bank sends payment instruction to clearing house, clearing house validates and routes to payee bank, payee bank confirms receipt and credits account, payee bank sends confirmation back through clearing house to payer bank. This entire flow takes seconds. The payer knows the payment succeeded before they close their phone. This is the core value proposition.

The message formats standardise how information is encoded: payer account, payee account, amount, currency, reference information, and optional extended data fields. The standardisation means any bank can send payments to any other bank without custom integrations. The network effects compound because the more banks in the network, the more value it provides to all participants.

Settlement Finality: The Irrevocability Problem

One feature that distinguishes RTP networks from older systems is settlement finality. In a card network, a transaction is authorised but not truly final until settlement batch runs. In RTP networks, the transaction is final at the moment of settlement, which is instantaneous. Once final, the payment cannot be reversed by the network. It can only be reversed if the receiving bank agrees to return the funds voluntarily.

This finality is the security advantage (no unauthorised changes) and the liability challenge (no automatic recourse if something goes wrong). Regulators are working to balance these: finality protects the integrity of the payment system, but the lack of recourse leaves customers vulnerable. The emerging consensus is that banks must absorb fraud losses up to certain thresholds and provide customer protection through their own fraud prevention systems, rather than relying on network-level reversal mechanisms.

---

Global Adoption Momentum and Next Frontiers

The momentum behind RTP adoption is accelerating globally. Over 70 countries have live networks or networks in development. The trajectory is clear: real-time payment infrastructure will be the standard in every developed economy within five years. Emerging markets are adopting even faster because RTPs solve problems that correspondent banking never could.

The next frontiers are cross-border interoperability and emerging market expansion. If payments between FedNow, SEPA Instant, UPI, and PIX could settle seamlessly, the global payment infrastructure would transform overnight. Banks and regulators are exploring this, but technical and regulatory obstacles remain. The world has not yet built a truly global RTP system that crosses borders and currencies seamlessly.