🇵🇰 Payments in South Asia outside India

Seven markets, seven architectures: mapping the landscape

South Asia outside India comprises seven payment markets with a combined population of nearly 600 million. Four of them (Pakistan, Bangladesh, Sri Lanka, and Nepal) built a national payment system in less than 15 years. Three smaller markets (Bhutan, the Maldives, and Afghanistan) round out the region, and their architecture is more unusual than that of many OECD countries. All seven share a common macroeconomic structure. They run foreign currency deficits, financed by the wages their migrant workers send home from the Gulf and Malaysia. No common currency, regional regulator, or shared scheme connects them. As a result, their central banks treat the payment system as a tool of monetary policy as much as a piece of infrastructure.

In this region, the price of acceptance is a policy decision. In its four large markets, administered pricing replaces commercial negotiation. In Sri Lanka, the central bank caps instant transfer fees by circular. In Bangladesh, it sets the maximum rate each category of provider may charge on an interoperable transfer. In Pakistan, the government subsidizes merchant QR payments to push cards out. A business model built on a Western-style percentage fee finds no pricing room at all in these four markets.

Pakistan: Raast, or how the state bypasses its own switches

Pakistan's retail payment infrastructure long rested on two private operators on top of a public settlement system. 1LINK, owned by a bank consortium since 1999, runs the national switch, which carries ATM, POS, IBFT interbank transfers, and bill payments. NIFT, a private joint venture founded in 1995, clears checks in three daily batches (regular, intercity, and same-day). Both settle their net positions in PRISM+, the State Bank of Pakistan's RTGS system, launched in 2008. The RTGS also serves as the central securities depository, an unusual combination. The system is now migrating to ISO 20022.

In 2021, the central bank launched Raast, a free public instant payment rail with alias-based addressing. The alias, called the Raast ID, is the user's mobile number. Raast bypasses 1LINK rather than sitting on top of it. A Raast transfer skips the private switch, pays no interchange, and settles directly in PRISM+. This setup makes the State Bank of Pakistan supervisor, infrastructure operator, and competitor to the firms it supervises all at once. That triple role shapes the entire Pakistani market.

In October 2025, the central bank announced its intention to route all government payments through Raast by the end of FY26. That covers public-sector salaries, pensions, social transfers, and government collections. Moving those flows builds a base of active accounts, and without that base the rail is of little use for private-sector acceptance. Brazil's and India's rails followed the same bootstrapping path: the government was the first customer before becoming the main driver of adoption.

Pakistan: PayPak, telecom wallets, and four ways into the market

PayPak, launched by 1LINK in April 2016, is the country's only domestic card scheme. It reports 16.1 million cards in circulation and about 28% of Pakistan's card market (1LINK, 2025–2026). Its pitch is cost. Scheme fees stay domestic, settlement is in rupees, and no foreign currency flows out to Visa or Mastercard. Its reach ends at the border. It offers no acceptance outside Pakistan. The card handles everyday in-store purchases and leaves travel and foreign online shopping to the international brands. In June 2026, 1LINK announced the country's first domestic credit card, run with Euronet Pakistan under a managed services model. The setup is meant to let mid-sized banks offer credit cards without building their own platform.

Unlike wallets in the European sense, the two leading wallets in Pakistan are banks. JazzCash (2012) is run by Mobilink Microfinance Bank, a subsidiary of the VEON/Jazz group. It reports about 21 million monthly active users, 57 million registered customers, and 850,000 merchants (company data, 2025). Easypaisa (2009), the pioneer of Pakistani mobile money, won SBP approval in 2025 to change its status. It became the country's first 100% digital retail bank. It has about 18 million monthly active users, roughly 14 million of them on the app (company data, 2025). Operating under a microfinance bank license, rather than an e-money license, is the defining feature of the Pakistani model. Any commercial negotiation is therefore with a licensed bank, subject to prudential rules and its own timelines.

Easypaisa's ownership structure is one of the items any 2026 due diligence has to cover. Easypaisa Digital Bank is 55% owned by Telenor Group and 45% by Ant Group, so Telenor holds control. The bank was carved out of the sale of Telenor Pakistan to PTCL, which closed at the end of 2025. In 2026, Telenor hired Citigroup to explore selling its stake. A change-of-control clause is how an acquiring or distribution contract deals with a sale of that stake during the contract term.

Bangladesh: where mobile financial services took over everything

In Bangladesh, mobile financial services (MFS) means a mobile payment service run by a commercial bank that holds an MFS license from Bangladesh Bank. The country follows a **bank-led model that few countries have kept for so long. The license goes to the bank, not to a telecom operator. The official register lists 14 MFS services at 13 banks**. Despite this apparent fragmentation, three brands hold most of the market.

Bangladesh's statistics show a wide gap between registered and active accounts. With 239 million MFS accounts for a population of about 175 million, duplicate accounts are rampant. One person commonly holds a bKash account, a Nagad account, and a bank account linked to a mobile banking service. Sizing a market on reported account numbers overstates the real addressable base by a factor of two to three. The only reliable metrics are transaction volumes and 30-day or 90-day active users, and the contract should require the partner to provide them.

The real economics of Bangladeshi MFS rest on *cash-out*, converting digital value received into cash. Users withdraw funds at an agent, who earns a commission on each withdrawal. That commission funds the network, which counts more than 1.8 million agents according to Bangladesh Bank. It also makes merchant payments structurally hard to push. A digital payment at a merchant replaces a withdrawal, and so wipes out the agent commission that came with it. In Bangladesh, the barrier to merchant payments is how the agent network gets paid, more than technical acceptance.

Bangladesh: public rails and interoperability by circular

Bangladesh Bank is one of the few central banks in the world, along with Bhutan's Royal Monetary Authority, to be the direct technical operator of its retail infrastructure. Since 2012 it has run the National Payment Switch Bangladesh (NPSB), which handles ATM, POS, online interbank transfers and, since November 2025, mandatory interoperability. It has run BEFTN since 2011, the country's ACH for credits and debits and the rail for salaries, pensions, and social transfers. It has also run BACPS since 2010 for check truncation and image clearing. Together, these two make up the Bangladesh Automated Clearing House. Settlement takes place in BD-RTGS (2015), a multicurrency system that handles the taka, the US dollar, the euro, the pound sterling, and the Canadian dollar. It therefore settles domestic interbank payments denominated in foreign currencies.

On October 13, 2025, the central bank issued a circular that reshaped the market. Since November 1, 2025, every bank, MFS operator, and PSP must route transfers from any account to any account through NPSB. The requirement covers bank-to-MFS, MFS-to-MFS, and PSP-related transfers. A bKash-to-Nagad transfer, which did not exist before, now falls under it. Beyond making the rail mandatory, the same circular caps the prices of these transfers.

The same gap shows up with Binimoy, the interoperable digital transaction platform (IDTP) billed at launch as “Bangladesh's UPI,” which links banks, MFS providers, and PSPs through aliases. Official half-year data shows the scale. It processed 217,213 transactions worth BDT 76.03 crore in the first half of 2025, or 0.00% of both volume and value across all the country's platforms. BEFTN alone processed 184 million over the same period. Binimoy works technically, but its share rounds to zero in national statistics. The central bank does not list it among its systems in operation. Calling it a national rail is not backed by any measured volume.

On the card side, TakaPay (2023) is Bangladesh's first domestic scheme, run by Bangladesh Bank through NPSB and launched with Sonali Bank, City Bank, and BRAC Bank. Its goal is explicitly macroeconomic: to reduce the foreign currency outflow paid to international schemes. A co-badge with RuPay has been announced for India–Bangladesh flows, which would be the region's first intraregional card bridge. So far, the co-badge is only an announcement, not a live service.

Sri Lanka: LankaPay and administered pricing for acceptance

Sri Lanka's payment infrastructure is run by a single company, LankaClear (Pvt) Ltd, jointly owned by the Central Bank of Sri Lanka, which supervises it, and the country's banks. LankaClear operates the entire national LankaPay system (Common Card and Payment Switch, CCAPS). It includes the CEFTS instant rail since 2015, the JustPay merchant overlay since 2016, the LANKAQR standard since 2018, and the domestic card scheme since 2019. The same company operates the legacy SLIPS ACH (1994) and CITS check image clearing (2006). Final settlement takes place in LankaSettle (2003), an RTGS system linked to the government securities depository.

Sri Lanka's shift to the instant rail was one of the fastest in the region. CEFTS grew from 8% to 67% of the country's interbank transactions in six years (CBSL / LankaPay). The shift came from a pricing policy imposed by circular, which made instant transfers cheaper than competing payment methods. Sri Lanka thus shows how a price cap can reshape the payment mix.

The National Card Scheme, launched in 2019 by LankaClear in partnership with JCB International, pairs a domestic scheme with an international acceptance network. Fees on domestic transactions are set locally and kept low, while cross-border acceptance runs on the Japanese partner's network. The scheme was accepted at more than 4,800 ATMs on the LankaPay network from day one (CBSL). The first phase covered debit. A second phase, marketed as “LankaPay 2 in 1,” adds credit. The arrangement gives a small market global reach without handing the economics of its domestic payments to an international scheme.

Sri Lanka is the region's most useful counterexample to the idea that mobile money wins everywhere. The central bank recognizes only two mobile payment systems in the country, both run by telecom operators. eZ Cash (Dialog Axiata PLC) is available without a bank account, including to customers of rival carriers. mCash (Mobitel, part of the Sri Lanka Telecom group, under the SLT-Mobitel brand) has a very large collection network covering bills, insurance, leasing, and agent banking. Both have connected to the national rails and to LANKAQR, but most Sri Lankan retail payments have gone through the CEFTS and JustPay bank rails, not through telecom wallets. The low cost and universal reach of the bank rails have left little room for telecom wallets.

GovPay, launched in 2025 by LankaClear on behalf of the government, is a one-stop platform for paying for public services, built on the LankaPay rails. It is a G2C/C2G setup: payments between government and citizens flow both ways over shared national infrastructure rather than through a private PSP. The same model exists, under other names, in the region's four large markets.

Nepal: a private QR network that became the de facto national standard

Unlike the rest of the region, Nepal's QR standard is run by a private operator. Everywhere else in South Asia, the QR standard is public: the central bank issues it, mandates it, and often runs it. In Nepal, a private company beat the public infrastructure. Fonepay, operated by Fonepay Payment Service Ltd (F1Soft group), handles more than nine out of 10 QR payments in the country (Nepal Rastra Bank / press reports). Public infrastructure exists alongside it. Nepal Clearing House Ltd (NCHL) runs the Retail Payment Switch and the NEPALPAY brand, both launched after the private network was already entrenched.

Nepal's public infrastructure has three layers. NCHL-IPS / NCHL-ECC (2012) is the clearing base, with check truncation and deferred transfers. Every bank integration in Nepal goes through it. connectIPS (2018) is the single real-time retail payment platform (web, mobile, API), run by the bank-owned clearing house. It is the default rail for government payments and school fees. Finally, NCHL's National Payment Switch / Retail Payment Switch carries alias-based payments (NEPALPAY Instant), request-to-pay, the interoperable QR, the biller gateway, wallet interoperability, and settlement for payment system operators. Members access all of it through an open API.

Nepal was also the first market where India's UPI was exported to a third-party QR network. Since March 2024, Indian travelers can pay Nepalese merchants with their UPI app by scanning a Fonepay QR code, under an agreement with NPCI International Payments Ltd (NIPL). More than 200,000 terminals are connected, according to NIPL (2025). In January 2025, NEPALPAY QR opened to visitors from China, Italy, South Korea, Malaysia, and Singapore. Descriptions of these agreements often blur three distinct things: a local merchant accepting a foreign QR code, two payment systems being interconnected, and a software stack being exported. Nepal is the first case. An Indian traveler can pay a Nepalese merchant, but a Nepalese resident has no equivalent route in India.

Migrant worker remittances reached NPR 1,723.27 billion in fiscal 2024/25, up 19.2% (Nepal Rastra Bank). They were equal to 26.5% of GDP in 2023, according to the World Bank, the highest dependence ratio in the region and one of the highest in the world. What a Nepalese payment rail needs to do therefore follows from processing inbound remittances and delivering them to recipients.

Bhutan, the Maldives, Afghanistan: three markets most overviews leave out

Bhutan, the Maldives, and Afghanistan make up the region's second tier, with far lower volumes than the big four. Each has an institutional setup that no large country offers in such a pure form. Bhutan entrusts its entire payment infrastructure to its central bank. The Maldives opened its instant rail to nonbanks from the design stage. In Afghanistan, the payment system has partly taken the place of the banking system.

Bhutan was the first country in the world to accept India's UPI for merchant payments, starting in 2021, under an agreement between NPCI International Payments Ltd and the Royal Monetary Authority. It later added domestic issuing. Four conditions came together: a small market next to India, a pegged currency, a central bank with sole decision-making power, and an integration managed end to end. Those conditions make Bhutan a benchmark case for studying how a payment standard is exported.

The three markets share one operational issue: concentrated counterparty risk. When the central bank is the sole switch operator, there is no commercial redundancy, no standard contractual remedy, and no alternative routing. An outage on the Bhutan Financial Switch or the Afghan ATS cannot be worked around through a private backup route. Elsewhere, an acquirer can hold a provider to its service-level commitments; here, no one owes them, because the switch operator is the central bank itself. Business continuity planning for a local acceptance contract in these three countries must therefore assume no contractual fallback.

Migrant remittances: the real engine, and what it means for anyone accepting payments

Migrant worker remittances are the funds that workers abroad send to their home countries. The region received about $200 billion in personal remittances in 2024 (World Bank). That money funds household spending and builds central banks' foreign exchange reserves, which explains its weight in payment regulation. The Gulf–South Asia corridor is the largest in the world by volume. Its two ends look very different. At the sending end, the flow is physical (exchange houses, cash, and migrant payrolls in the UAE, Saudi Arabia, Qatar, and Kuwait). At the receiving end, it is digital (wallets, bank accounts, mobile money). The corridor's cost structure comes from that asymmetry, not from the technology used.

South Asia is the cheapest region in the world to receive a remittance, even after prices rose in 2025. Those low prices come from competition among Gulf exchange houses, the corridor's sheer volume, and government incentives that subsidize the formal channel. A new entrant therefore finds no pricing margin to capture on inbound remittances. Whatever revenue these flows generate comes from downstream transactions, once the money has been credited to the recipient.

*Hawala is a network of independent brokers (hawaladars) who settle with one another on trust, with no cross-border movement of funds for each transaction. The FATF and the IMF classify it as an informal value transfer system (IVTS). It is faster, cheaper, and often the only channel available for Afghanistan and the region's rural areas. Dubai is its main hub. Its Bangladeshi form, hundi*, is the explicit target of the government's 2.5% incentive. Hawala is the benchmark against which every formal operator in the region is measured on price and speed.

• Check the currency regime before technical integration. The Pakistani rupee, the taka, the Sri Lankan rupee, and the Nepalese rupee are under exchange controls: repatriating funds from a local merchant account is a regulated transaction. It is the first item in the contract, not the last.
• Never size a market on reported accounts. Bangladesh has 239 million MFS accounts for ~175 million people: ask for 30-day or 90-day active users and volumes, not sign-ups.
• Check the counterparty's status in the regulator's register, including its exact phase: Pakistan's EMI register distinguishes in-principle approval, pilot, and commercial operations; Nepal's register gives the actual legal name (IME Khalti Ltd., Esewa Ltd.); the MFS entity in Bangladesh's register sometimes differs from the brand (Trust And Pay Ltd.).
• Model the end of subsidies. The Raast P2M subsidy runs until June 30, 2026, with a PKR 3.5 billion budget. A Pakistani acquiring model that does not survive its expiry is not a business model.
• Treat price caps as an input, not a risk. LKR 25 per CEFTS transfer, 0.15%/0.20%/0.85% by category in Bangladesh: these limits are not negotiable, you live with them, and they decide which business model works.
• Distinguish foreign QR acceptance, system interconnection, and software stack exports. An Indian traveler paying through Fonepay in Nepal or LANKAQR in Sri Lanka does not mean the two systems are interconnected, or that payments flow the other way.
• Audit business continuity where the central bank is the sole operator (Bhutan, Bangladesh, Afghanistan): there is no commercial redundancy, no standard contractual remedy, and no alternative routing.