Nepal's Digital Payment Revolution
Nepal went from cheque-heavy retail payments to over a trillion rupees a month on real-time rails in a few years. How the shared-infrastructure model worked, and what is left to solve.
By Velloreth
Nepal’s digital payment story is usually told through the wallet apps — eSewa, Khalti by IME, Namaste Pay — because those are the brands people touch. The more important story is underneath them: a small country built shared payment infrastructure that most richer markets have not managed, and the volume now moving through it has crossed from “growing fast” into “this is the system of record.” By mid-2026, Nepal’s Retail Payment Switch was processing more than NPR 1.2 trillion in a single month, and its annual real-time volume had surpassed the cheque system it was designed to supplement.
This post walks through how that happened — the rails, the wallets, the merchants, and the cross-border turn — and what remains unfinished.
From cheques to switches
Not long ago, retail payments in Nepal meant cash, cheques, and a slow clearing process. Electronic fund transfers existed, but the everyday experience of moving money between people and businesses was dominated by physical instruments and branch hours.
The institutional answer was the Nepal Clearing House Limited (NCHL), promoted by Nepal Rastra Bank, which operates the country’s core payment infrastructure: electronic cheque clearing (NCHL-ECC), the Interbank Payment System (NCHL-IPS) for net settlement between institutions, CORPORATEPAY for bulk and business payments, connectIPS for real-time retail payments, and — critically — the National Payment Switch (NPS), which includes the Retail Payment Switch (RPS), the National Payment Interface (NPI), and NEPALPAY QR.
The design philosophy is the part worth understanding. Rather than letting each bank and wallet build a private network — the siloed model that kept payments fragmented in many markets — Nepal’s regulator and NCHL built common rails with rules. Wallets and mobile banking apps connect to the same switch; a payment from any wallet can reach any participating bank or merchant. Interoperability was an infrastructure decision made before the market got locked into silos, and it is the single biggest reason the system works the way it does.
The scale of that system, per NCHL’s own reporting: over 211 million transactions worth more than NPR 18.6 trillion processed across its platforms in one recent fiscal year — roughly 67% growth in volume year over year — accounting for the large majority of digital retail settlement value in the country.
The Retail Payment Switch milestone
The Retail Payment Switch, launched in late 2021, is the real-time engine underneath everyday payments: account-to-account transfers, interoperable QR payments, wallet and mobile-banking initiated transactions through the NPI, merchant payments, bill payments, and request-to-pay services.
The trajectory has been steep. NCHL reported that RPS processed roughly NPR 9.3 trillion of real-time transaction value across fiscal year 2025/26 — having crossed the annual value of cheque transactions during the same period — and set a monthly record above NPR 1.2 trillion in mid-2026. Those are not fintech-pitch numbers; they are national-infrastructure numbers, reported by the operator under regulatory oversight.
Alongside volume came product maturity. NCHL began commercial rollout of the NPS-NCS and NEPALPAY Card — a domestic card scheme with data-localization emphasis — and has been layering tokenization onto connectIPS for e-mandates, wallet loads, and e-commerce checkouts, replacing raw account details with revocable tokens. Request-to-pay via NEPALPAY Request and virtual payment address support round out a rail set that increasingly looks like a full modern payment stack rather than a single switch with bolt-ons.
The wallet layer people actually use
On top of the rails sits the consumer-facing market: eSewa (from the F1Soft group, Nepal’s first wallet and a licensed payment service provider), Khalti — which merged with IME Digital in 2025 to form IME Khalti, consolidating two major networks — plus Namaste Pay, CG Pay, Hamro Pay, and others licensed under NRB’s payment framework.
Each wallet competes on user experience, agent networks, and merchant offers. None of them competes on the fundamental ability to move money to any bank or merchant — because that ability comes from the shared rails, not from each wallet’s private infrastructure. This is the structural advantage Nepal has over markets where wallet interoperability arrived late or never: a user switching wallets does not lose their ability to pay; a merchant accepting one wallet accepts effectively all of them through NEPALPAY QR and the NPI.
The consolidation into fewer, larger licensed players — NRB has actively pushed PSP consolidation — mirrors what other markets found: payment services carry heavy compliance and infrastructure costs, and scale matters. For consumers, the practical effect is a market of a few serious wallet options rather than dozens of half-maintained apps.
QR: the acceptance revolution
The visible face of the transformation is the QR sticker. Merchant acceptance of QR payments scaled from a few hundred thousand points in the early 2020s to over 2.3 million by early 2025 — driven by interoperable NEPALPAY QR, zero or near-zero acceptance costs for small merchants, and the simple economics that a printed sticker beats a card terminal on every axis that matters to a kirana store.
The pattern is familiar from India and should be: where merchant acceptance costs collapse to effectively zero, digital payment volume follows into the informal economy. Nepal’s QR growth did the same thing UPI’s QR layer did — moved the cash-versus-digital decision from a margin calculation for the vendor to a default for the customer. When a shop has a QR sticker, paying by phone is easier than counting notes.
The fee structure supports it: connectIPS transaction fees for fund transfers and merchant payments have sat in the low single-digit rupees per transaction across small-value slabs — a cost a merchant can absorb on thin margins in a way that a percentage-based card MDR never allowed. Clearing services have also benefited from VAT-exempt treatment under recent finance legislation, keeping the cost stack low.
Government, bills, and the boring categories
The less-glamorous categories did as much for adoption as person-to-person transfers. Government revenue payments, utility bills, mobile top-ups, credit card payments, insurance premiums, education fees, and capital-market transactions all route through NCHL’s platforms — connectIPS handles government payments at scale, and revenue systems like RMIS integrate with the rails for real-time and near-real-time collection.
This matters for adoption economics in a way consumer apps rarely match: a citizen who pays their electricity bill and their child’s school fees digitally every month has permanently crossed the behavioral gap. The wallet is then used for merchant payments almost as an afterthought. Government digitization was the onboarding engine; consumer apps were the engagement layer.
The regulator’s hand
It would be a mistake to attribute the transformation to entrepreneurs alone. Nepal Rastra Bank has been an active architectural participant: licensing the payment service provider category that wallets operate under, directing consolidation among PSPs to reduce fragmentation, publishing annual payment systems oversight reports that make the sector’s growth measurable, and establishing a Digital Finance Innovation Hub in 2025 to engage with new models on a structured basis.
The oversight reports matter more than their readership suggests. By tracking transaction volumes and values across every rail — RTGS, IPS, connectIPS, RPS, mobile banking, wallets, QR — NRB produces the only complete picture of where money in Nepal actually moves. That visibility is what lets regulation respond to reality rather than to anecdotes; it is also what gives international observers (development banks, regional payment operators, correspondent institutions) confidence in the system’s trajectory.
The consolidation push is the other notable policy thread. A market of dozens of marginal PSPs cannot sustain the compliance, security, and infrastructure investment that a national payment system requires. NRB’s direction toward fewer, better-capitalized licensed players — visible in the IME-Khalti merger and the dismissal of non-compliant licenses — trades short-term variety for long-term stability. Consumers lose some choice; the system gains operators who can afford to invest in tokenization, fraud detection, and uptime.
This regulator-as-architect model — common rails, licensed operators, published oversight, active consolidation — is the closest thing Nepal has to a national industrial policy for digital services, and it is worth studying precisely because it was executed by a small central bank rather than a rich one.
The India connection
The most consequential recent development is outward. UPI QR acceptance for Indian travelers has been live at Nepali merchants on the Fonepay network — retail, hospitality, pilgrimage sites — reducing the cash friction for the hundreds of thousands of Indian visitors Nepal hosts annually. For a tourism economy, that is direct economic plumbing, not a tech demo.
Then, in June 2026, the linkage deepened: India’s UPI and Nepal’s NPI were connected for real-time person-to-person cross-border transfers, implemented by NPCI International and NCHL. Indian and Nepali users can send money across the border using mobile numbers or virtual payment addresses through participating banks — no account numbers shared, no multi-day wire process, settlement in near real time.
For a remittance corridor of Nepal’s economic importance, replacing a chunk of slow, fee-laden transfer channels with a rail-to-rail linkage is structural. It also completes an architectural arc: Nepal’s shared-rail model gave it something worth connecting to, and the connection now makes the domestic infrastructure part of a regional network rather than a national silo.
What is still unfinished
The revolution is real; the completion is not. Four gaps remain material.
Fraud and dispute resolution on push payments. Real-time rails are hard to reverse — a lesson every market learns the hard way. Social engineering, mule accounts, and misdirected transfers put pressure on wallet operators and banks to build better fraud detection and clearer dispute processes. Tokenization and risk-based limits help at the margins; the fundamental tension between instant settlement and consumer recourse is a policy problem, not a switch problem.
Digital literacy at the bottom of the pyramid. The urban, phone-literate population is well served. Older users, rural users, and first-time account holders interact with the same apps without the mental models to spot scams or recover from mistakes. The infrastructure assumed literate users; inclusion now requires the harder work of meeting people where they are.
Rural merchant and user coverage. QR acceptance is dense in the Kathmandu valley and urban centers; it thins rapidly outside them. The agent networks that wallets built help, but the last-mile of digital payments — like the last mile of broadband — follows roads, grid power, and commercial density.
Merchant economics at scale. The low-fee model that drove adoption has to be sustained by someone. As volumes mature and the infrastructure costs are amortized, the industry and the regulator will negotiate who pays for what — the same zero-MDR sustainability debate India has been having for years. Getting this wrong in either direction (fees that kill adoption, or costs that hollow out the operators) has visible consequences; getting it right is the quiet work of the next decade.
The shape of what was built
Step back and the achievement is architectural more than technological. Nepal — landlocked, smaller in population than most of its neighbors’ cities, with a GDP that could not fund a greenfield national payments program à la India’s UPI at the same scale — built a shared-rail system that interoperates, settled at national volume, hosted a competitive wallet market on top, crossed the trillion-rupee monthly mark, and linked cross-border to the largest real-time payment system in the world.
The lesson for peer markets is the one the infrastructure itself embodies: interoperability as public infrastructure, decided early, enforced by a regulator willing to make incumbents share the rails, beats private-network competition for everyone except the would-be monopolists. Nepal made that choice before the market calcified. The volumes now moving through the switch are the dividend.