Fintech Weekly Deep Dive — UPI’s Free Lunch Era Nears Its End | Week of July 20–26, 2026

Executive Summary

India’s government is actively considering reintroducing a Merchant Discount Rate (MDR) on UPI transactions for large merchants — the most consequential policy shift in digital payments since zero-MDR was mandated in January 2020. Under the proposal reported by the Economic Times on July 16, merchants with annual turnover of ₹1–1.5 crore would pay 5–7 basis points on UPI transactions exceeding ₹2,000. Person-to-person transfers and payments to small merchants would remain untouched.

This is not a sudden move. It is the culmination of a growing consensus — spanning the Parliamentary Standing Committee on Finance, the Department of Financial Services, the Payments Council of India, CareEdge Research, and senior bankers — that UPI’s zero-fee model has become structurally unsustainable at its current scale. The numbers are stark: UPI processed 241.6 billion transactions worth ₹314 lakh crore ($3.7 trillion) in FY 2025-26, yet the government’s subsidy for the ecosystem covers barely 11% of what the industry says it costs to run. The proposed MDR on large merchants would generate an estimated ₹3,500–5,000 crore annually — significantly more than the ₹2,000 crore allocated in Budget 2026.

For consumers, the immediate impact is negligible. The levy targets 2–4% of merchants — large retailers, e-commerce platforms, and organised chains — not the neighbourhood kirana or individual users. But the long-term implications are profound: this is India acknowledging that even digital public infrastructure needs a business model, and that the transition from adoption-driven subsidies to sustainability-driven economics has begun.

The Story in Depth

Context: How India Built the World’s Largest Free Payment System

When the government mandated zero MDR on UPI and RuPay debit cards in January 2020, the intent was unambiguous: remove every friction point between cash and digital. UPI was still processing 12.5 billion annual transactions then — impressive for a four-year-old system, but a fraction of what was to come. Zero-MDR was the rocket fuel.

And it worked. UPI’s transaction volume has grown nearly 12,000-fold since its April 2016 launch, from 373,000 transactions in its first month to 66 crore daily in FY26. ¹ The platform now accounts for 85% of India’s digital payment volume, 49% of global real-time payment transactions, and is operational across 12 countries. ² Over 700 banks are live on UPI, up from 21 at launch. The person-to-merchant (P2M) segment alone accounts for 62% of volume, with over 700 million QR codes deployed across kirana stores, fuel stations, and e-commerce platforms. ³

The zero-MDR policy served its purpose spectacularly. India went from a cash-dominant economy to one where digital modes account for 99.8% of transaction volume and 93% of value in retail payments. ⁴ But nothing in economics is truly free — and the bill has been quietly accumulating.

What Happened This Week

The July 16 Economic Times report — citing three people aware of the matter — revealed that the government is examining a proposal to levy MDR on UPI transactions for merchants with annual turnover exceeding ₹1–1.5 crore, and only for transactions above ₹2,000. The proposed rate of 5–7 basis points (0.05–0.07%) would mean a merchant paying ₹5–7 on a ₹10,000 UPI payment. ⁵

This was not an isolated leak. It followed a cascade of signals:

• March 2026: The Parliamentary Standing Committee on Finance, in a report dated March 12, called for MDR restoration, stating that “the sustained expansion of UPI requires a viable revenue mechanism” and that “the absence of MDR makes the UPI ecosystem financially unsustainable.” ⁶
• February 2026: Budget 2026 allocated ₹2,000 crore for the Digital Payments Incentive Scheme — an increase from the ₹1,923 crore spent in FY25, but far below the industry’s demand for ₹4,500 crore, and a fraction of the estimated ₹5,800 crore monthly cost of operating UPI. ^{7 8}
• Ongoing: The Payments Council of India (PCI), representing 180 non-banking payment companies, has made multiple formal representations requesting a “controlled MDR” of 0.3% on large-merchant UPI transactions. ⁹

The government has not yet taken a final decision. The proposal awaits inter-ministerial deliberation and potential public consultation. But the direction of travel is unmistakable.

Separately this week, NPCI continued expanding UPI’s capability envelope. The July 23–24 briefs revealed that NPCI is developing an offline NFC-based “Tap & Pay” feature allowing UPI transactions up to ₹2,000 without internet connectivity — targeting transit hubs, underground metro stations, and rural areas with poor connectivity. ¹⁰ UPI also clocked record numbers in March 2026: 22.6 billion transactions worth ₹29.53 lakh crore in a single month. ¹¹

The juxtaposition is telling: even as UPI adds new capabilities and breaks volume records, the question of who pays for all of this is becoming impossible to ignore.

Why It Matters

The subsidy math doesn’t work anymore. In FY 2021-22, the government provided ₹1,300 crore in UPI subsidies. By FY 2025-26, the estimated annual cost of running UPI exceeded ₹69,600 crore (₹5,800 crore monthly), while the government subsidy sat at ₹2,000 crore — covering roughly 2.9% of the system’s operating cost. ^{8 12} Industry experts estimate banks incur 3–4 basis points per UPI transaction in direct processing costs, including NPCI switching fees, technology infrastructure, and reconciliation overheads. ¹³ At 241 billion transactions, that translates to roughly ₹20,000–25,000 crore annually just in bank processing costs — before accounting for fintech app development, merchant onboarding, customer support, and fraud prevention.

The government’s position, articulated by RBI Governor Shaktikanta Das, has been that UPI is “important infrastructure” that the government has “consciously chosen to keep free for users by subsidising it.” ¹⁴ But subsidies are fiscal policy, not infrastructure policy. Every rupee spent subsidising UPI is a rupee not spent elsewhere. As UPI’s scale has made it too big to fail, it has also made it too big to subsidise indefinitely.

The proposed MDR is surgically targeted, but the signal matters more than the rate. At 5–7 basis points affecting only 2–4% of merchants, the immediate revenue impact is estimated at ₹3,500–5,000 crore annually — modest relative to UPI’s total throughput. ⁵ But this is the thin edge of the wedge. Once the principle of MDR on UPI is established, the rate and scope can (and likely will) be adjusted over time. The Payments Council of India originally proposed 0.3% (30 basis points) — nearly 5–6 times the current proposal. ⁹ The parliamentary committee’s language was even more open-ended, calling broadly for a “viable revenue mechanism.” ⁶

For the payments industry, this is a lifeline. Zeta’s Mehul Mistry framed it precisely: for banks, PSPs, and payment infrastructure providers who have “absorbed processing costs for years without direct monetisation,” this could “unlock a sustainable revenue model.” ⁵ The current model forces payment companies to treat UPI as a loss-leader, cross-subsidised by credit card MDR, lending products, or venture capital. That works when you’re a well-funded startup chasing growth. It doesn’t work when you’re a bank processing 66 crore transactions daily and watching your infrastructure costs climb.

Data & Metrics

The subsidy collapse is the hidden story. The government paid ₹1,300 crore in UPI subsidies in FY 2021-22. By FY 2024-25, this had fallen to ₹1,923 crore — a number that looks like an increase until you realise transaction volumes nearly doubled over the same period. In per-transaction terms, the subsidy has fallen from roughly 19 paise per transaction in FY22 to about 8 paise in FY25. The Banker reported that subsidies fell to as low as ₹200 crore in FY25 before being revised upward. ¹² The ₹2,000 crore allocation for FY27, while nominally higher, represents a continued decline in real terms.

Expert Views

Mehul Mistry, SVP at Zeta: “This could be the most consequential shift in India’s digital payments landscape since the zero-MDR mandate in 2020. While the proposed 5–7 bps MDR would apply to only 2–4% of merchants, those above ₹1–1.5 crore in annual turnover, this segment drives an outsized share of UPI’s merchant transaction value.” ⁵

Parliamentary Standing Committee on Finance (March 2026): “The committee would like to emphasise that establishing a viable revenue mechanism is critical to ensuring the UPI ecosystem achieves financial sustainability without perpetually straining the government exchequer.” ⁶

Department of Financial Services (to the Standing Committee): “The objective of introduction of zero MDR was to make digital transactions affordable and widely accessible. However, the absence of MDR makes the UPI ecosystem financially unsustainable.” ⁶

Nikhil Pahwa, Founder of MediaNama (2020, prescient): “By removing MDR, the Government of India has effectively destabilised a well-established market norm of incentives in digital payments, and will end up creating perverse incentives in the payments ecosystem.” ⁶

CareEdge Research (October 2025): “Zero MDR on UPI [is] challenging sustainability of system.” The report noted that while UPI is well-positioned as the backbone of India’s payments landscape, its zero-fee framework creates long-term structural risks. ¹⁵

Tamal Bandyopadhyay (via LinkedIn): “The zero-MDR regime onboarded 500 million users, but government incentives cover only 11% of industry costs.” ⁸

RBI Governor Shaktikanta Das: Described UPI as “an important infrastructure” that the government has “consciously chosen to keep free for users by subsidising it — a policy that has borne rich dividends in terms of adoption.” ¹⁴

Consumer Impact

Immediate impact: Zero for most users. If implemented as proposed, the MDR would apply only to merchants with annual turnover above ₹1–1.5 crore, and only on transactions exceeding ₹2,000. This means:

• Person-to-person transfers: Completely unaffected. Your ₹500 split at a restaurant, your monthly rent payment, your salary credit — all unchanged.
• Kirana stores and small merchants: Completely unaffected. The neighbourhood provision store doing ₹50 lakh in annual turnover pays nothing.
• Large e-commerce platforms and organised retail: This is where the levy bites. Amazon, Flipkart, BigBasket, Zomato, Swiggy, and similar platforms would pay 5–7 paise per ₹100 on UPI transactions above ₹2,000.

The passthrough question. The critical consumer concern is whether large merchants will absorb this cost or pass it through as checkout fees, surcharges, or price increases. At 5–7 basis points, the per-transaction cost is small — ₹5–7 on a ₹10,000 order. For most consumers, this would be imperceptible. But it sets a precedent. If the MDR expands in scope or rate over time (and historical patterns suggest it will), the passthrough risk grows.

There is also a broader behavioural risk. Experts quoted by Inkl warned that “even a modest MDR could discourage acceptance of UPI for low-value transactions, slowing the pace of digital adoption” among merchants near the threshold. ¹⁶ A kirana store doing ₹90 lakh in turnover today might deliberately suppress reported turnover to stay below the ₹1 crore threshold, or might start steering customers toward cash for transactions just above ₹2,000. These are marginal effects, but India’s digital payments story has been built on marginal effects compounding at scale.

The positive consumer angle. A sustainable UPI ecosystem means continued investment in reliability, security, and innovation. The NPCI’s offline tap-to-pay feature, AI-based fraud detection, cross-border expansion, and credit-on-UPI all require funded infrastructure. The ₹2,000 crore government subsidy is demonstrably insufficient for this. An MDR-funded model, even one starting at 5–7 basis points, creates a revenue stream directly tied to usage — aligning incentives between payment providers and the merchants and consumers they serve.

Looking Ahead

Watch for the formal proposal. The government has not yet issued a draft notification or opened public consultation. The next signals will come from either an inter-ministerial meeting or a formal draft from the Department of Financial Services. Given the Parliamentary committee’s March recommendation and the ET report, a formal proposal before the end of 2026 appears likely.

Watch the threshold and rate. The ₹1–1.5 crore turnover threshold and ₹2,000 transaction floor are politically calibrated to exclude small businesses. But the 5–7 bps rate is a starting position. The PCI’s 30 bps proposal and the committee’s open-ended language suggest the industry will push for higher rates once the principle is established.

Watch for payment method steering. If MDR returns only for UPI while cards remain at existing MDR levels (1.5–2.9% for credit cards, 0.4–0.9% for debit cards), large merchants may start incentivising UPI less aggressively or routing high-value transactions through other rails. Conversely, if UPI MDR remains far below card MDR, the shift toward UPI at the merchant layer will only accelerate.

Watch the global context. Brazil’s Pix system — UPI’s closest global peer — has also faced MDR debates as it scaled. The July 22 brief noted growing US-Brazil tension over Pix’s geopolitical influence. ¹⁷ India’s MDR decision will be watched by every country building or scaling instant payment systems. The question isn’t just whether India can make UPI financially sustainable — it’s whether it can do so without undermining the inclusion gains that made UPI a global model.

Sources

Covering developments from July 20–26, 2026. Published by CashlessConsumer.