DPI Deep Dive — Tuesday | October 20, 2026

Focus layer: L2 — Payments & Financial Rails (UPI, NPCI and adjacent bank rails)
Coverage period: 13–19 October 2026; information checked through 08:30 IST on 20 October

Executive Summary

This week’s payment story is not a new UPI feature but a postponed price change. On 13 October, NPCI moved the planned merchant discount rate (MDR) on selected person-to-merchant payments to “immediately post Diwali”, without naming a date. The delay came just before a reported hotel-association protest, while NPCI’s own explanation cited the festive season. The fee is a merchant-side charge under the announced framework, not a charge on customers; but merchant responses can still affect checkout choices. Daily UPI figures show extraordinary throughput, while the government’s new financial-inclusion camp plan connects payment access to identity, banking, insurance and fraud awareness. The open test is implementation: clear rules, usable redress and better public measures of quality.

Key Developments

1. The MDR start date moved; the framework did not disappear

On 13 October, NPCI’s UPI and Services Steering Committee decided that MDR on “select” person-to-merchant (P2M) UPI transactions would begin immediately after Diwali, on a date to be notified separately. That superseded the previously scheduled 15 October start. The NPCI notice gives one reason—“keeping in mind the festive season”—and no replacement date, implementation circular or explanation of which operational steps remained incomplete. This is a deferral, not evidence that the framework has been withdrawn.

The government’s 15 September explanation describes the framework’s boundaries. It says person-to-person payments remain free regardless of amount; merchant transactions up to ₹2,000 remain free; and small merchants receiving up to ₹1 lakh a month through UPI QR under the P2PM category are protected by zero MDR. The announced standard rate is 0.4% on eligible merchant transactions above ₹2,000, capped at ₹300 for transactions of ₹75,000 and above. The release also lists different treatment for specified essential sectors and capital-market payments. Those details matter: “UPI charges” is a misleading shorthand for a set of selected merchant-side rates and exemptions.

The same official explanation says banks should ensure merchants do not pass the charge to customers and bars UPI apps from imposing platform fees or hidden charges. Consumers therefore should not be told that an MDR line item is a government levy or a required UPI customer fee. Yet a formal prohibition on passing through a fee cannot guarantee that all retail behaviour stays unchanged. Merchants can choose which payment methods to accept, and operational confusion can itself become checkout friction. Any new start date should arrive with a single, public rulebook: which merchant categories qualify, how threshold and exemptions are calculated, how disputes are handled, and what a customer should do if asked to pay an extra fee.

For the ecosystem, the question is not merely whether a fee exists. MDR revenue is meant to help fund payment-system participants, but the policy also needs to preserve the low-friction acceptance that made UPI useful. A date-only announcement leaves merchants, banks, app providers and consumers unable to prepare consistently. The next notification should say whether the original terms remain intact, how settlement and refunds will work, and whether the transition is staged.

2. Merchant resistance made the policy’s distributional question visible

The timing of the postponement followed a concrete public warning from one trade group. On 13 October, the Kerala Hotel and Restaurant Association said its members would stop accepting UPI payments above ₹2,000 from 15 October in protest. Livemint reported the announcement the next day, alongside NPCI’s deferral. This is evidence of a stated plan by an association, not proof that a statewide boycott was carried out. Nor does it prove that the protest caused NPCI’s decision: the official notice cites the festive season, while press reports described merchant representations as part of the context.

Still, the episode shows why “merchant pays” does not end the consumer discussion. A merchant-side fee has no automatic right to become a customer surcharge under the government’s stated safeguards. But merchants can change their acceptance behaviour, and the reported proposal to refuse larger UPI payments would have shifted inconvenience to customers even without adding a fee to the bill. For a customer who relies on a phone-based bank account, that is a practical access question, not just an accounting distinction. For a hotel, restaurant or other high-turnover business, a fee on selected high-value transactions can accumulate differently from a fee on an occasional purchase.

The proposed framework is also unusually sensitive to classification. A ₹2,000 transaction threshold, a monthly-receipt test for a small merchant, sector-specific flat charges and special capital-market rates require reliable merchant categories and transaction metadata. If an establishment is misclassified, the burden can appear as an unexplained deduction. If an exemption is difficult to claim, the nominal protection may not reach the smallest shops. A serious rollout should publish an accessible eligibility checker, sample settlement statements, a correction path and a clear complaint route before charging starts.

Public policy should not assume either that every merchant will absorb the fee or that every merchant will retaliate by refusing UPI. Those are outcomes to measure. When the revised date is announced, regulators and NPCI should monitor acceptance rates, merchant complaints, reported surcharges and failed or abandoned checkouts by merchant class. Otherwise, “no customer charge” will remain a formal assurance with no public evidence about how the policy behaves at the point of sale.

3. Six days of transaction cards show scale, not reliability

NPCI’s daily statistics cards available by this morning report between 828.00 million and 843.29 million UPI transactions per day from 13 to 18 October. Summed across those six days, the cards record 5,013.77 million transactions worth ₹593,564.42 crore—about 5.014 billion payments and ₹5.94 lakh crore. The day-by-day sequence is 833.98 million on 13 October; 835.76 million on 14 October; 828.00 million on 15 October; 834.64 million on 16 October; 843.29 million on 17 October; and 838.10 million on 18 October. The latest card available at 08:30 IST on 20 October covers 18 October; a figure for 19 October had not yet appeared in the daily series at the time of writing.

The cards’ totals exclude credit-card-on-UPI and credit-line-on-UPI transactions, as NPCI notes. They are useful evidence of payment scale, but not an audit of success or welfare. They do not show attempted transactions that failed, how quickly a debited user received a reversal, fraud loss, outage duration, merchant-level MDR exposure or variation by bank and region. A high count can coexist with difficult dispute resolution; a lower daily value does not by itself identify why the value changed. The data also cannot validate the government’s estimate that about 4% of merchant transactions would attract MDR: the cards do not break out eligible merchant transactions or exemption status.

This measurement gap matters to both sides of the pricing debate. If the stated purpose of MDR is to sustain the payments ecosystem, public evidence should include the costs being covered and what the collections finance. If the policy says most transactions remain free, people need a transparent denominator and a way to test who qualifies. NPCI’s volume series is a start, not a substitute for acceptance, reliability and cost data.

The adjacent rails underline the need to compare like with like. The RBI’s RTGS FAQ, updated 14 October, describes RTGS as a large-value system with a ₹2 lakh minimum, no RBI-set upper amount ceiling and 24×7 availability; it also sets bank-charge caps. RTGS is not a like-for-like QR checkout replacement for UPI. But the FAQ is a useful reminder that India’s payment stack has multiple rails with distinct purposes, customer protections and pricing rules. A good public dashboard would report those differences rather than letting one headline transaction count stand in for the whole system.

4. Financial-inclusion camps join identity work to payment access

A second in-window development comes from the Department of Financial Services. Its campaign announcement sets a period of 21 October to 31 December and a stated reach of 2.70 lakh Gram Panchayats and Urban Local Bodies. Bank camps are planned for PMJDY account opening, Aadhaar seeding for PMJJBY and PMSBY, re-KYC, pending C-KYC IDs and nominations, insurance and pension enrolment, Kisan Credit Card and crop-insurance applications. The programme also includes digital-fraud awareness and outreach about unclaimed financial assets.

This is a planned service push, not evidence that 2.70 lakh camps have already taken place or that applications have been completed. Its connection to payments is foundational: a bank account, correctly linked identity where required, working contact details and a usable grievance path all influence whether someone can receive and use digital transfers. It also crosses three DPI layers at once—identity, financial rails and trust. Pairing identity-linked tasks with financial products makes clear explanation of purpose, consent and correction procedures especially important. A camp should not be measured only by forms collected or accounts opened.

The useful public scorecard would show planned and completed camp locations, attendance, service requests completed on the first visit, pending cases and reasons, re-visit rates, complaints and time to resolution. It should disaggregate by service and location without exposing people’s personal data. Fraud-awareness sessions should include practical recovery steps and contact routes, not only generic warnings. If residents leave with an unresolved Aadhaar mismatch or a failed seeding request, the event has moved a queue rather than solved a problem.

The campaign may improve the conditions for digital payment use, but account inclusion is not the same as active, safe access. A working rail depends on both the network and the human support around it. The announced camps are an opportunity to measure that support locally—and to make the difference between onboarding numbers and an actually usable service visible.

Trends & Implications

  • A delayed date is a governance problem, not merely a calendar edit. NPCI has not cancelled the framework, but the gap between the original start and the new post-Diwali window leaves merchants without a precise implementation schedule. The full circular should precede enforcement.
  • Fee incidence and user experience can diverge. The announced safeguards say MDR belongs to the merchant ecosystem, yet merchant acceptance choices can still affect customers. Monitoring must cover both surcharges and refusal or fallback behaviour.
  • Scale needs a reliability companion. Six days exceeded five billion reported transactions, but the series lacks failures, reversals, fraud and distribution data. Public infrastructure deserves public performance measures.
  • Digital inclusion is cross-layer work. Camps linking accounts, identity, insurance and fraud literacy can support payment access, provided participants get a clear purpose, usable support and a remedy when links fail.

Upcoming Milestones

  1. NPCI’s exact post-Diwali implementation date and operational circular, including category definitions, exemptions, caps and complaint handling.
  2. Evidence of whether merchant groups’ announced acceptance changes were carried out, and any subsequent policy engagement.
  3. A public measure of MDR collections, eligible transaction counts, allocation to payment providers and the small-merchant support fund.
  4. A campaign dashboard after 21 October showing camp coverage and completed versus pending services, with reasons for failure and grievance resolution.
  5. NPCI payment statistics that pair transaction volume and value with failure, reversal, fraud and service-availability measures.

Sources