Scaling UPI: The Economics Behind NPCI’s New Operational Limits

Policy Update
Manish Prakash Shinde

Background

Although the Unified Payments Interface (UPI) has been operational for a decade, it does not exhibit the obsolescence typical of legacy infrastructure; rather, it currently faces challenges associated with rapid and unprecedented scalability. NPCI launched it in April 2016 to make bank-to-bank transfers instant and interoperable, so a payment from any bank app could reach any other bank account without the usual friction. The decision that really made it explode came later: in January 2020, the government set UPI’s Merchant Discount Rate to zero, so neither payer nor payee would ever see a fee. That single choice is arguably why the neighbourhood vegetable seller and the metro-station tea stall started accepting UPI as readily as a large retail chain.

The numbers since then are hard to absorb fully. UPI’s transaction value has gone from about ₹21.3 lakh crore in FY2019-20 to over ₹213.8 lakh crore by January 2025, a roughly tenfold jump in five years. Annual volumes climbed from 118 billion transactions in 2023 to 172 billion in 2024, and in FY2025 alone, close to 145 billion transactions worth over ₹220 trillion moved through the network. Somewhere in that curve, UPI stopped being a convenient app feature and became national infrastructure, the kind other countries now study when building their own digital public rails. However, infrastructure at that scale runs into questions nobody had to answer when volumes were smaller: who pays to keep the servers running, who governs the handful of apps carrying most of the traffic, and what happens when the system strains under its own popularity. Since August 2025,
Since August 2025, NPCI, the RBI and the government have begun answering those questions through four connected moves: technical limits on how the UPI network itself can be used, a long-pending cap on how much volume any single app can carry, a stricter authentication rule for every transaction, and, as of this month, the first piece of legislation that could eventually allow a charge on large merchants. This update walks through each of the four in turn, and the shared economic logic that connects them the cost of running a payments network that has always been free at the point of use.

NPCI’s approach works by rate-limiting the APIs banks and PSPs use to talk to the UPI network, rather than by restricting transactions directly. The ten APIs capped from August 2025 covering balance enquiries, autopay mandate processing, and transaction-status checks are all requests that consume server capacity without ever completing a payment. Throttling their frequency reduces load on the network while leaving actual money movement untouched. Banks and PSPs that exceed the caps face API restrictions or a freeze on onboarding new customers, which places the enforcement cost on the entities generating the traffic rather than on end users.

Functioning

Under the RBI’s Authentication Mechanisms for Digital Payment Transactions Directions, 2025, every UPI transaction now needs two distinct authentication factors drawn from something the user knows (a PIN), something they have (a registered device or token), or something they are (biometrics), and at least one of those two has to be dynamic, generated fresh for that specific transaction rather than reused. Entities that do not comply are on the hook for fraud losses on those transactions. It is one more step in the flow, in exchange for a payment system that’s harder to compromise with a stolen static credential.

There is an older, slower-moving rule sitting alongside all this. Since November 2020, NPCI has sought to prevent any single UPI app from carrying more than 30 per cent of the total transaction volume, measured over a rolling three-month window, with a warning at 25–27 per cent. In theory, an app over the cap must stop onboarding new users. In practice, the deadline keeps moving, now pushed to December 31, 2026, the second extension since the rule was proposed, because PhonePe and Google Pay, between them, still handle more than 80 per cent of UPI’s volume, PhonePe alone close to half. It is worth asking whether a rule can be said to exist at all if it has never once been enforced against the players for whom it was written. Smaller PSPs, meanwhile, say the compliance burden of monitoring and throttling their own API traffic falls hardest on those with the least engineering capacity to spare.

Performance

Somebody has to pay for “free,” and that somebody has mostly been the government, through the Incentive Scheme for Promotion of RuPay Debit Cards and Low-Value BHIM-UPI Transactions, run by the Department of Financial Services, which reimburses banks and PSPs for part of what it costs them to process fee-free transactions. The allocations over the last few budget cycles show the strain: ₹2,484.97 crore actually spent in FY2023-24, a Budget Estimate for FY2025-26 of just ₹437 crore, sandwiched between an FY2024-25 allocation revised upward mid-year from ₹1,441 crore to ₹2,000 crore, and an FY2026-27 Budget that put the figure back up to roughly ₹2,000 crore.

Source: Compiled from Union Budget documents, cited in IndianStartupNews (2025) and YourStory (2025); PIB (2025) for the ₹1,500 crore FY2024-25 Cabinet-approved P2M scheme.

That kind of swing is not just messy accounting; it reflects how far the subsidy has fallen behind the real cost of running the network. Estimates floated around the FY2024 Budget put the annual cost of processing UPI merchant transactions above ₹2,000, a category that received zero incentive cover at close to ₹8,000 crore, with roughly ₹9,000 crore needed to compensate the ecosystem across all UPI merchant transactions fully. Even the Cabinet-approved ₹1,500 crore scheme for FY2024-25 covered only transactions up to ₹2,000 to small merchants with a turnover under ₹20 lakh, at 0.15 per cent per transaction. Which leaves an obvious gap: nothing in the current scheme covers the higher-value merchant transactions that the government is now, separately, considering a charge on.

Impact

Operationally, the implementation of transaction limits has yielded a direct, positive impact on infrastructure resilience. By curbing non-financial traffic—specifically redundant status pings and repeated balance checks—the National Payments Corporation of India (NPCI) has successfully mitigated server load constraints. The primary impact of this optimization is the stabilization of the network; it ensures high system uptime and reduces technical decline rates for genuine financial transactions, even as overall network volume continues to scale.

Conversely, assessing the economic impact of the proposed Merchant Discount Rate (MDR) remains speculative, as the legislative framework enacted this month merely lifts the legal prohibition on merchant fees without establishing concrete rates. Regulatory authorities, including the Finance Ministry and the Reserve Bank of India (RBI), have clarified that peer-to-peer transfers and low-value consumer payments will remain exempt. Therefore, the prospective impact is structurally confined: any eventual fee would exclusively target high-value, large-merchant transactions exceeding ₹2,000.

From a systemic impact perspective, this targeted approach is highly significant. Estimates indicate that while this segment constitutes only 4% of UPI’s transaction volume, it represents approximately 67% of its total transaction value. Consequently, introducing a modest MDR on this narrow segment would have a dual impact: it could generate substantial revenue to close the infrastructure funding gap for Payment Service Providers (PSPs) and banks, while simultaneously shielding the broader retail consumer base from new costs. This would provide the necessary capital for continuous technological investment to maintain network resilience.

However, the potential behavioral impact on the market must be evaluated against historical precedents. The 2017 revision of card MDR rules, which tiered charges by merchant turnover, resulted in significant pushback from the retail sector. Even if the macroeconomic rationale for a large-merchant MDR is sound, its implementation risks negative downstream impacts. Large retailers may attempt to pass these new operational costs onto consumers, or they may alter their payment acceptance behaviors for high-ticket transactions. Ultimately, the success of the policy will depend not only on resolving the funding gap but on mitigating the adverse financial impact on the merchant ecosystem.

Emerging Issues

Funding that swings wildly from year to year: the incentive scheme has moved from ₹2,485 crore to ₹437 crore to ₹2,000 crore across three budget cycles, making it hard for banks and PSPs to plan around. Suggestion (industry bodies): a multi-year, formula-linked funding commitment instead of a fresh discretionary allocation every year.
A market-cap rule that keeps getting postponed: the 30 per cent TPAP cap has been deferred twice since 2020 and will not come due till December 2026, while PhonePe and Google Pay’s combined share has stayed above 80 per cent throughout. Suggestion (competition and fintech analysts): use the extension window to grow bank-owned UPI apps actively, UPI Lite, and UPI 123PAY, rather than waiting out the clock again.
Compliance costs that fall unevenly: smaller PSPs say real-time monitoring of their own API traffic takes engineering resources that larger players have. They do not (PSPs): shared compliance tooling from NPCI, or timelines phased by provider size.
A messaging gap: recurring rumours that “UPI is about to start charging users” have forced the Finance Ministry into repeated, reactive denials, most recently this week. Suggestion (government): one standing, regularly updated public FAQ, rather than clarifications issued only after speculation spreads.
No finalised criteria yet: reported turnover thresholds for a future MDR range from ₹1.5 crore to ₹4 crore, and rate estimates from 0.04 to 0.4 per cent. Suggestion (merchant associations): publish a draft framework and open it to stakeholder comment before any Section 10A notification is issued.

Way Forward

Synthesizing these regulatory and structural shifts reveals an evolving governance paradigm for the Unified Payments Interface (UPI). To ensure long-term ecosystem viability without compromising the core value of frictionless, zero-fee transactions for retail users, regulators must implement the following targeted solutions:

Tiered Monetization Framework: The Reserve Bank of India (RBI) and NPCI should formalize a targeted Merchant Discount Rate (MDR) exclusively on high-value corporate transactions (e.g., above ₹2,000). A portion of this revenue must be earmarked for a dedicated infrastructure fund to subsidize core banking server upgrades and maintain systemic resilience.
Milestone-Based Regulation: Instead of repeatedly deferring market-share caps, regulators must transition to a predictable compliance roadmap. Implementing dynamic API-level routing can naturally redistribute transaction flows to emerging market players without imposing hard caps that degrade the user experience.
Technical Efficiency Protocols: To mitigate server congestion during peak hours, operational guidelines must strictly throttle non-financial network traffic, such as redundant balance inquiries. Furthermore, aligning security with risk profiles using passive authentication for micro-transactions and strict multi-factor authentication for high-value transfers will optimize processing loads.
Proactive Stakeholder Governance: Policymakers must replace reactive announcements with a codified operational framework. Mandating advance public disclosures and formal economic impact assessments prior to implementing fee or API changes will stabilize market expectations and prevent commercial disruptions.

Ultimately, transitioning UPI from a phase of rapid user acquisition to structural maturity requires actionable, transparent governance. By balancing targeted monetization with continuous infrastructure investment, regulators can ensure that UPI’s next decade remains as financially sustainable as its initial growth was transformative.

Selected References and Important Links

Business Standard. (2026). Govt clarifies UPI to remain free; MDR may apply to large merchants later. Business Standard. https://www.business-standard.com/

Business Today. (2026). BT Exclusive: Large merchants may bear UPI MDR under proposed framework, say sources. Business Today. https://www.businesstoday.in/

Business Today. (2026). UPI charges explained: PCI says consumers, small merchants will continue to pay nothing amid MDR debate. Business Today. https://www.businesstoday.in/

Entrackr. (2026). Exclusive: NPCI mandates new MCC for UPI gift cards, caps per-transaction limit at Rs 10,000. Entrackr. https://entrackr.com/exclusive/exclusive-npci-mandates-new-mcc-for-upi-gift-cards-caps-per-transaction-limit-at-rs-10000-11810736

MediaNama. (2025). NPCI extends UPI market share cap deadline to 2026, Google, PhonePe relief. MediaNama. https://www.medianama.com/

Ministry of Finance, Government of India. (2025). Cabinet approves Incentive scheme for promotion of low-value BHIM-UPI transactions (P2M). Press Information Bureau. https://pib.gov.in/

Policy Circle. (2026). UPI MDR needs a shield for small merchants. Policy Circle. https://www.policycircle.org/policy/upi-mdr-small-merchants/

Prime Minister’s Office, Government of India. (2025). Cabinet approves Incentive scheme for promotion of low-value BHIM-UPI transactions (P2M). Government of India. https://www.pmindia.gov.in/

Razorpay. (2026). UPI Transaction Limit Per Day in 2026: Bank-Wise Daily UPI Limits. Razorpay Blog. https://razorpay.com/blog/

The Week. (2026). Will UPI payments be charged? Govt clarifies amid speculation over MDR rates. The Week. https://www.theweek.in/

Zee News. (2025). From August 1, these UPI transactions will have limits under NPCI’s new API rules. Zee News. https://zeenews.india.com/

About the Contributor

Manish Prakash Shinde is a Research and Editorial Intern at IMPRI and a final-year Bachelor’s student in statistics at Fergusson College, Pune. His academic interests lie in Behavioural Economics, Corporate Governance & ESG.

Acknowledgement

The author extends sincere thanks to the IMPRI team for their guidance.

Reviewed by: Vishal Kumar and Himanshi Singh

Disclaimer:

All views expressed in the article belong solely to the author and not necessarily to the organisation.

Reviewed by: Vishal Kumar & Himanshi Singh