UPI Won the Payment War. Now It Has a Revenue Problem.

Everyone loves the UPI story. And why wouldn’t they? A country that had 190 million unbanked adults a decade ago now runs the world’s largest real-time payments network. The Unified Payments Interface processed 228.3 billion transactions worth USD 3.4 trillion in 2025. It handles roughly half of all real-time payment volume on the planet. It has displaced Visa in global transaction count. It is, by any metric, one of the most extraordinary infrastructure achievements in financial history.

But here is the thing nobody says loudly enough: UPI generates almost nothing.

Not for the banks processing the transactions. Not for the platforms like PhonePe and Google Pay that built hundreds-of-millions-of-user businesses on top of it. Not for NPCI, which operates the whole system as a not-for-profit and survives on the thinnest of surpluses — INR 1,552 crore in FY2025, in a network processing USD 3 trillion. The government compensates payment processors with an annual subsidy of INR 1,500 crore in lieu of merchant fees. The industry estimates that a 0.1% MDR — one-tenth of one percent — on UPI’s FY2025 value would generate USD 3.1 billion in annual revenue. Revenue that, under current policy, simply does not exist.

India built the world’s most impressive payment network and then legislated the business model out of existence.

I have spent months analyzing the data for our 2026 India Fintech report, and the MDR question is the single most consequential policy variable in the entire sector. Whether it gets reinstated — or doesn’t — determines the terminal value of every payment company in India. And yet it barely gets mentioned in the breathless coverage of India’s fintech boom.

Let me be precise about what zero-MDR actually means in practice.

PhonePe has 657 million registered users. Forty-seven million merchant partners. A 45% share of UPI volume. It is India’s most valuable private technology company, targeting a USD 15 billion IPO. Its payment revenue from all of those transactions? Effectively zero. The adjusted profit of INR 630 crore that PhonePe reported for FY2025 — its first year of profitability — came almost entirely from financial services attached to those payment relationships. Insurance. Lending. Revenue from the payment itself: negligible.

This is not a criticism of PhonePe. It is a structural description of the market they operate in. And it is the reason PhonePe’s financial services revenue grew 206% year-over-year in FY2025, while its core payment revenue grew at a fraction of that rate. The company has understood the game correctly: payments are customer acquisition. Revenue comes from the financial products sold to those customers afterward.

The same logic applies across the board. Google Pay stays in UPI because it wants Indian consumer data and distribution. Paytm is recovering from its regulatory near-death experience by doubling down on its 1.17 crore Soundbox merchant device base — not because the devices generate meaningful direct revenue, but because they anchor merchant relationships for lending, insurance, and cash management products. Even CRED, whose 13 million users generate an ARPU approximately 20 times that of mass-market UPI apps, built its business model around using credit card payment management as the hook for a premium financial services flywheel.

Everyone knows the payments layer is the loss leader. Everyone knows the money is upstream. The question is whether the government knows this creates a structural problem — and whether it will act before the problem becomes systemic.

Here is the uncomfortable scenario.

India’s digital payments market will more than double to approximately USD 958 billion by 2030. UPI will process something in the range of USD 10 trillion annually by FY2030, according to PwC projections. If MDR remains zero throughout that period, Indian payment companies will have built one of the world’s largest financial infrastructure platforms — and extracted essentially no direct economic value from it.

What they will have extracted is data. And data, in India’s fintech ecosystem, is being systematically converted into credit, insurance, and investment products. The Account Aggregator framework — which facilitated USD 19.5 billion in loans in FY2024–25 via consent-based data sharing — is the mechanism. The ONDC network. The Unified Lending Interface. India Stack, taken as a whole, is an extraordinary machine for turning payment flows into financial product distribution.

So perhaps zero-MDR is, in fact, the right policy. Force the infrastructure layer to be free, and let competitive advantage accumulate at the product layer. It is not entirely unlike what happened with the internet — the pipes were commoditized, the application layer extracted all the value.

But I am not sure that framing holds, for one important reason: sustainability.

The government subsidy covering processor costs is rising — INR 1,500 crore in FY2024–25, INR 2,000 crore budgeted for FY2026–27. It will need to keep rising as volumes grow. At some point, the question of who pays for the rails becomes unavoidable. Banks have been quietly lobbying for MDR reinstatement for years. The payments industry argument — that even 0.1% MDR on P2M transactions would transform the economics of the sector without materially affecting consumer adoption — is technically sound.

The counter-argument is political: digital payments are a flagship government achievement, zero-fee UPI is something millions of users actively value, and any policy reversal carries real electoral risk. So the government keeps delaying, subsidizing, and hoping the financial services monetization layer develops fast enough to make the question moot.

That bet may pay off. PhonePe’s 206% financial services revenue growth suggests the monetization thesis is real. But it concentrates value at the few platforms with the scale and product breadth to execute the conversion from payment utility to financial services platform. For everyone else — the smaller payment apps, the bank-issued UPI handles, the niche players — zero-MDR means building on ground that will never be economically solid.

The UPI story is genuinely remarkable. I believe that sincerely, and the data in our report backs every superlative. But remarkable infrastructure with a broken revenue model is a problem that compounds over time, not one that resolves itself. India has built the world’s best payment rails. The next decade will reveal whether it also built a financially sustainable ecosystem on top of them.

The question for investors is not whether UPI works. It clearly does. The question is whether the handful of platforms converting UPI relationships into credit and insurance revenue will be enough to justify the astronomical capital that has been, and continues to be, deployed in Indian payments.

My read: probably yes, for the top two or three platforms. For everyone else, the math remains very hard.