Why brick and mortar will never bank Pakistan's other 100 million

A road trip to Southern Punjab, and the idea that became EasyPaisa

The episode opens with Muzamil framing the file before the guest answers a single question. Pakistan has been told for a decade that fintech is about to change the very dynamic of how finance is done. The cash in circulation is still extraordinarily high. The update every year is that there is a big problem to solve and the hope is to solve it — and yet it remains unsolved. He brings in Irfan Wahab Khan, chairman of EasyPaisa, to find out where it actually stands.

Irfan’s macro framing is direct. “Today even half of the population is unbanked or underbanked, which is a huge huge challenge for us,” he says. “If we talk about 100 million people, 100 million people who don’t have a bank account. How can we bring our countrymen into a financial inclusion net? And I am very clear that from traditional banking it is not possible. From brick and mortar it can never happen, because we have not achieved that in seventy years.”

The origin story is more granular than the corporate version. Irfan was the first employee at Telenor Pakistan in 2004. By the mid-2000s, on a road trip from Karachi to Peshawar during the wheat harvest in Southern Punjab, he met an elderly farmer whose son was sending money home from Saudi Arabia. The man had no bank account and had to travel to the nearest town every time money arrived. “Then I thought,” Irfan says, “why can’t we use this tool that every person has in their hand?” He read the M-Pesa case study from Kenya, went back to the Telenor board, and was turned down. The board reminded him that Telenor was a telco, that reputation risk on money was huge, and that the company had no expertise. He kept pushing. In 2009 the State Bank issued the branchless banking licence. That was the birth of EasyPaisa as a concept.

What “digital bank” actually changes — and what microfinance could never do

Muzamil presses on the licence question early, because it is the structural piece most listeners will not understand. EasyPaisa already had a microfinance bank and a wallet. So what changes now that the regulator has issued a digital retail banking licence?

Irfan’s answer is the cleanest list of the conversation. Microfinance came with caps — on lending sizes, on the kinds of products that could be offered, on cross-border activity. “Cross-border payments you cannot do directly. Your licence does not allow you, because your purpose is something else.” Under the digital retail bank, all of that opens up. Business accounts for SMEs. Car loans. Home loans. Mortgages. Much larger ticket sizes on personal lending. Direct wallet-to-wallet cross-border transfers. Buy-now-pay-later. And — the product Muzamil deliberately stays on — credit cards.

The transition path is also unusual. “We are going to be transitioning or transforming our microfinance licence into the digital retail bank,” Irfan says. “This is the first case of that, because we were already there compared to the newcomers who will just get the new digital retail licence.”

The scale Muzamil pulls out of him is what gives the licence its weight. Fifty million accounts — “every fourth Pakistani is an EasyPaisa customer.” Over 16 million active wallets transacting on a regular basis. Around five million transactions a day. Over 600,000 loans disbursed every day. And the throughput figure that lands hardest: the platform processed transactions worth roughly nine percent of Pakistan’s GDP in 2024, against a cash-in-circulation pile of around nine trillion rupees — roughly twice the level of comparable economies.

The Alipay turn, and why blockchain came in through remittances

The architectural pivot that made the scale possible came when Irfan, by then Telenor Pakistan CEO, took two decisions. First, buy back the microfinance partner and take full control of the bank. Second, find a global partner who could complement Telenor’s strengths in distribution and customer insight with technology and fintech depth.

That partner was Ant Group. Ant came in for a 45% stake against Telenor’s 55%, in what Irfan describes as the foundation of the modern EasyPaisa. The product implications followed quickly. Nano-lending scaled. A new app launched in 2019. And — the most technically interesting use case Irfan describes — EasyPaisa piloted blockchain on international remittance, opening a corridor between Malaysia and Pakistan because Telenor’s Malaysian operation could close the loop on the other side. “We were able to establish and show it to the regulators that with technology you can do twenty-four-seven transactions without any fee at the best rate,” he says. The pilot did not become a product because the microfinance licence did not allow cross-border activity. The digital banking licence now does.

The unbanked customer is not the upper-middle-class wallet user

Muzamil draws the cleanest line in the conversation between two very different EasyPaisa customer profiles. For an upper-middle-class user, the wallet is “a fancy little tool” — a nice-to-have add-on. For someone who has never had a bank account, the digital bank is the first banking relationship of their life. He recalls that growing up, conventional banks judged customers by their dressing before opening an account.

Irfan’s response leans on data and AI. Even a household maali or driver has a profile the platform can build — what the salary inflow looks like, what mobile top-ups they buy, what their remittance pattern is. “On the basis of that you can actually draw a profile, and on the basis of that profile you can decide that if he needs a loan, up to what amount, and as he pays back, his credit history builds up and you keep on increasing the ticket size.”

The same logic extends to insurance and savings. Irfan flags the EasyPaisa OPD medical insurance launched during COVID, and an insurance campaign targeting workers physically installing billboards. On the savings side he uses the Chinese reference deliberately — Alipay’s leftover-change money-market fund became the largest in the world by surpassing JPMorgan’s, not by chasing big tickets but by letting a hundred rupees sit overnight and earn something. EasyPaisa is now, by his account, Pakistan’s largest mutual-fund distributor by number of customers — not by ticket size — because the minimum entry is a thousand rupees rather than fifty or a hundred thousand.

Why Pakistan missed the credit-scoring boat

Muzamil’s sharpest provocation in the middle of the conversation is on credit scoring. Pakistani banks, he argues, have been structurally lazy — content to lend to the government at risk-free returns while private credit lags. The data exists in the market. Food Panda knows what someone is spending. Daraz knows what they are buying and whether they paid cash on delivery or returned the order. None of it gets stitched into a usable national risk profile.

Irfan does not push back. “We missed the boat because we were too closed-loop as banks. We did not share data. Credit bureaus’ reach was also limited because your traditional banking customers are very few. If you have only one or two million customers in total, you cannot score twenty or twenty million people on that base.”

The remedy he points at is structural rather than commercial. NADRA, he notes, was ahead of the world fifteen years ago — but the mindset that data must not be shared, especially with the private sector, has not changed. He contrasts this with India’s Aadhaar APIs, which opened the identity layer to any startup that wanted to build on it. “Data should not be seen as a one-time money-making opportunity,” he says — meaning the regulator should not see open APIs as a single-transaction revenue stream. It is the asset that the next generation of innovation will be built on.

EasyPaisa’s own AI credit-scoring engine, developed with a partner and trained on the platform’s transaction history, is in production for internal lending. Irfan confirms the company is now considering opening it up as a service to selected partners — with the explicit hope that the partner data flowing back enriches the model further.

API rails, SMEs, and the integration problem fintechs keep failing on

Later in the discussion Muzamil pulls on a thread that anyone who has tried to launch a small e-commerce store in Pakistan will recognise. EasyPaisa’s checkout on Daraz is seamless. As an SME on Shopify, integrating a payment gateway is, in his words, a nightmare. The big players get prioritised because they bring volume. The small players sit in a long queue. Fintechs historically build for the former and ignore the latter.

Irfan does not contest the observation. “Your observation is absolutely correct,” he says, and then names the three causes: internal bureaucracy, regulatory KYC and documentation requirements, and a prioritisation logic that follows volume. The fix, in his framing, has two parts. The first is competition — more digital banking licences mean more players chasing the same SME segment, which forces simplification. The second is regulatory: State Bank has to make KYC and onboarding lighter at the bottom of the market.

The conversation pivots from there into the One QR question. Muzamil is sharp about the international examples — India’s roadside fruit cart with a printed QR, China’s tier-three city tea stall accepting payment by phone. Pakistan’s One QR initiative, he points out, has been talked about for years. Where is it?

Irfan’s answer is the most candid piece of policy commentary in the episode. The State Bank, he says, has done a lot — Raast is a real piece of infrastructure, and the digital retail licences are forward-looking. But: “We need to really make it fee-free. So that you can roll it out, and then on the backend the rails that are used are Raast. It should not be that — okay, I rolled this out so the incentive comes to me, and that becomes a disincentive for anyone else to use it. You cannot expect a retailer to have multiple QRs. It does not work like that. It is cumbersome. It is complex. It will not fly.” Single QR, free at the acquirer layer, Raast underneath — that is the only configuration he believes scales.

Women, gold, and the inclusion gap behind the GDP gap

The most consequential macro point Irfan makes lands almost in passing. Pakistan’s female labour-force participation, he notes, sits at roughly 18% against a 50% population share. That gap, more than any other single variable, is why Thailand, Vietnam and the Philippines — economies that were smaller than Pakistan twenty or thirty years ago — are now meaningfully ahead. “Everyone is contributing towards economic growth, and the key for that is inclusion. Financial and digital inclusion.”

EasyPaisa’s product response, he says, has been to lean into accounts held in women’s sole names, particularly in conservative districts where the historical default has been a husband’s or brother’s account. And to extend nano-loans to women running home-based businesses — bakers, small retailers — on the strength of their own profile rather than a male guarantor. The point is not a CSR talking point. It is the bottleneck on the country’s growth model.

What digital banks should — and should not — build first

Muzamil pushes Irfan on what he considers a gap in the Pakistani digital banking offering: budgeting tools, transaction tagging, AI-powered financial assistants, voice-driven interfaces — the layer that wraps a bank account in something closer to a financial buddy. He argues the utility is highest precisely for users with the lowest financial literacy.

Irfan agrees on budgeting — pilots are running in-app and will surface soon. He is more cautious on investment advice. “That comes with a lot of responsibility. Outside, the accredited investor concept exists. You have to fill up a profile, answer questions about your experience, your understanding. So we have to be careful from a reputation viewpoint, because you want to give advice and tomorrow there is an issue.” His framing is segment-by-segment: advice rolls out to qualified, aware customers first, then broadens as the regulatory and trust groundwork settles.

The phase-one focus, he is clear, is different. Get awareness of what a digital bank actually does. Build the trust that lets a customer treat the wallet as their prime account, not a side pocket holding five or seven thousand rupees. Once salary lands in the account, savings, investing, bill payment, transfers and e-commerce checkout all sit one tap away.

Credit cards, BNPL and cross-border wallets — the next-product list

On credit cards Irfan is clearer than at any other point in the conversation. “You will see very soon our offering on the credit cards as well.” Loan limits, similarly, are no longer capped by the licence — they are capped only by what the data says about a given customer. BNPL is on the same shortlist, explicitly framed against a credit-card-like instalment structure, sized by the AI engine.

On remittances, the Alipay+ partnership opens the second leg. A Pakistani traveller in the Far East paying a hawker through their EasyPaisa wallet. A Pakistani expat in Dubai sending money home wallet-to-wallet, arriving in real time, with no fee. A freelancer in Pakistan repatriating dollar earnings from a foreign wallet. “That will open up another layer of opportunity actually for cross-border payments,” Irfan says.

The network as public infrastructure — BISP, floods, charity rails

Muzamil draws out one strength of EasyPaisa that does not show up in the fintech narrative: the retailer and distribution network, which has been used as quasi-public infrastructure for fifteen years. EasyPaisa partnered with the Benazir Income Support Programme as far back as 2009 — government-to-people welfare payments routed through a private rail because the rail reached further than any bank branch could. The same network has been used for flood relief disbursements and NGO cash programmes.

The charity rail is the one Irfan is proudest of. EasyPaisa users can contribute as little as a hundred rupees a month on recurring basis. Billions of rupees move through this channel annually — small tickets, large aggregate, all of it through a digital interface that did not exist a decade ago.

Pakistan in 2050: leapfrog, tokenisation and the no-legacy advantage

By the end of the conversation Muzamil zooms out. If banks become instrumental in identifying the grey economy, putting it on the books and then optimising it — as a previous guest, Faisal Aftab of Zayn Capital, argued — what does Pakistan look like in 2050?

Irfan’s answer is the one he is most willing to commit to on the record. “This is one area where we can leapfrog.” The barrier of trust around sending money digitally, he says, has been broken. What remains is the right inflection point, the right use cases, and the global rails — tokenisation, fractional ownership, regulated crypto, central bank digital currencies — that the rest of the world is now building out in parallel.

The strategic point underneath it is the one most Pakistani technology arguments miss. “We don’t have to worry about the legacy systems. We don’t have anything. We are starting fresh.” A young population, native digital citizens, no entrenched magnetic-stripe-and-cheque infrastructure to protect. The same condition that has kept Pakistan poor on the bank-branch metric for seventy years is, on the digital-rails metric, an advantage.

Muzamil closes by saying out loud what he says he tells every fintech founder he meets: he wants 2025 to be the year of the inflection point. He acknowledges the limitations Irfan has spent the conversation walking through — the regulator’s pace, the data-sharing impasse, the SME integration friction, the cash habit. And he asks the question that every viewer was going to ask anyway: have you actually used EasyPaisa, and if not, why not?