Media365
India has done something remarkable with UPI. We did not merely digitise an existing payment instrument. We created an open, interoperable, account-to-account payment infrastructure at population scale. But the more fundamental idea behind UPI was strategic: To build a home-grown payment rail that would reduce our dependence on international card schemes and their economics.
India could have simply accepted the existing card architecture and its pricing model. Instead, we built our own. An editorial in this paper (“A small fee, a stronger UPI”, IE, September 17) has welcomed the decision to levy a fee. But I would ask a simple question: If India could innovate so dramatically on payment technology, why should we not innovate on pricing too?
UPI is fundamentally a credit-transfer system. The payer initiates a transfer from their bank account to the beneficiary’s account. RTGS, NEFT and IMPS follow the same basic principle.
Cards are economically different. A card transaction is essentially a pull transaction: The merchant initiates the transaction and the customer’s account or credit facility is debited based on authorisation.
The question is not whether UPI should have a sustainable economic model. It should. The question is whether a payment rail deliberately designed as an alternative to card networks should automatically adopt a percentage-based pricing logic developed for card networks. We engineered our way out of dependence on the old payment rails. Can we also price our own rail on our own terms?
India’s credit-transfer systems have traditionally used simple, fixed or slab-based pricing, with maximum charges, rather than percentage-of-value pricing. NEFT historically had maximum charges of Rs 5 for transfers up to Rs1 lakh and Rs 25 above Rs 1 lakh. RTGS charges were capped at Rs 25 for Rs 2–5 lakh and Rs 50 above Rs 5 lakh.
And importantly, IMPS — itself an NPCI-developed, real-time credit-transfer product — followed the same broad philosophy. In 2016, the government directed public-sector banks that maximum charges for IMPS transactions above Rs 1,000 should not exceed the corresponding NEFT charges.
When India created its own real-time credit-transfer system, it did not build a percentage-of-value pricing model. It used simple, low, slab-based pricing with maximum charges. So why should UPI now require 0.40 per cent MDR, with a maximum of Rs 300?
UPI operates at an extraordinary scale. An IIM Bangalore analysis of NPCI’s 2024–25 financials calculated total NPCI expenses of Rs 2,270 crore against 230.2 billion transactions — approximately 9.8 paise per transaction across NPCI’s payment systems. On the narrower operational-cost basis we have been examining, excluding marketing, the figure is around 5 paise per transaction. The wider ecosystem obviously has other costs — banks, merchant acquirers, fraud management, security and customer support. But the question remains: How much revenue is genuinely required for cost recovery and sustainability? NPCI also demonstrates what scale and frugality can achieve. There is perhaps something for the wider ecosystem to learn from that.
UPI and India’s wider Digital Public Infrastructure were not created merely to maximise transactions or create revenue pools. They were built to expand economic participation. A small merchant accepting Rs 50, a migrant transferring money instantly, or a citizen accessing a digital public service are not merely transactions. They represent participation in the formal economy. Banks and technology companies need sustainable economics. The answer is not to reject markets or profit. But sustainability should not make us lose sight of the purpose for which this infrastructure was created.
Markets can create wealth. Businesses can make profits. Technology can create extraordinary efficiency. But economic efficiency must ultimately ensure well-being. That is why we built UPI as DPI. Technology is the means. Inclusion and human capability are the purpose. Perhaps the next innovation in UPI should therefore be as bold as the first: Not simply how we move money, but how we price the infrastructure that moves it.
The writer is former chief operating officer, NPCI and formerly with the World Bank – Payment System Development Group
