Under the September 15 National Payments Corporation of India circular on Unified Payments Interface transactions, of the 0.4% Merchant Discount Rate collected from the merchant by the acquiring bank, 0.28% is paid as interchange to the issuing bank. The issuing bank pays 0.12% as a Payment Service Provider fee to the payer-side PSP bank, which pays 0.08% as an app-provider fee to the Third-Party Application Provider.

The MDR is distributed across four participants. The issuing bank retains 0.16%, or 40% of the total; the acquiring bank keeps 0.12%, or 30%; the Third-Party Application Provider receives 0.08%, or 20%; and the Payment Service Provider bank retains 0.04%, or 10%.

Three observations follow. First, the issuing bank—the customer’s bank—is the single largest beneficiary, retaining 40% of MDR at 0.16%. This is the same institution that already benefited from the widened spread between the Standing Deposit Facility rate and savings bank deposit rates throughout the zero-MDR period—SBI cut its savings bank rate to 2.70% in 2020 and held it there through the entire 2022–24 rate-hike cycle, even as the SDF rate rose to 6.25%. It now receives the largest commercial share of the restored MDR in addition.

Second, the TPAP ecosystem—PhonePe, Google Pay, and Paytm—receives 20% as a direct revenue line for the first time, a long-overdue recognition for platforms that bore losses throughout the zero-MDR period.

Third, where a merchant is acquired by a Payment Aggregator—Razorpay, Cashfree, or PayU—the acquiring bank’s 0.12% is split with the PA under their bilateral commercial arrangement. The PA’s effective revenue is not specified in the circular and will vary by sponsor bank.

The Authority Question
NPCI is a Section 8 not-for-profit entity owned by a consortium of banks. The PSS Act vests regulatory authority in the Reserve Bank of India. The August 2026 amendment removed the statutory bar on charges; it did not explicitly delegate pricing authority to NPCI.

A private banking consortium that simultaneously sets the MDR rate, decides its distribution among participants, and charges its own switching fee outside that distribution—with all three decisions sitting in one organisation—is a governance arrangement that warrants public examination. That NPCI is also the largest undisclosed beneficiary of the scheme through its switching fee adds to the concern.

The Small-Merchant Paradox
Large merchants—e-commerce platforms, major retail chains, and corporate enterprises—have the commercial sophistication to negotiate their effective MDR rates, threaten alternative channels, or restructure payment flows. Small P2M merchants—the kirana that just crossed the P2PM threshold, or the mid-size restaurant—have none of that leverage. The merchant that most needs protection from an inequitable design is the merchant least equipped to negotiate around it.

Point 10 of the circular proposes a small-merchant fund: a 5% contribution from each participant’s MDR share, directed towards merchants with annual turnover of ₹2 million or below. The intent is sound, but the fund is not yet operative—its framework will be finalised with the RBI within three months.

More significantly, the fund’s eligibility criterion is annual business turnover of ₹2 million—the established RBI definition—while the MDR exemption uses the narrower P2PM criterion of ₹100,000 per month in UPI receipts. NPCI applies two different definitions of a small merchant simultaneously: the narrower one to deny the exemption, the broader one to channel fund support.

A merchant with ₹1.8 million in annual turnover receiving ₹150,000 per month through UPI fails the exemption and qualifies for the fund—paying MDR with one hand, and receiving a fraction of its proceeds with the other.

The Missing Dimension
FAQ’s forty-two questions cover sustainability, competition, consumer protection, and category-specific rates. Not one addresses UPI fraud, victim compensation, or a risk pool.

NPCI’s chief executive stated in April 2025 that reported UPI fraud is below one basis point of transaction value. That figure is accurate. It is also measured on a perimeter—the Fraud Monitoring Return, covering bank-reported fraud above ₹100,000—that severely undercounts authorised push-payment fraud, where victims report to the police rather than their bank.

The Parliamentary Standing Committee on Communications and IT found that approximately 10.4% of reported fraud is recovered. Ninety paise in every rupee is permanently lost to the victim.

The Reserve Bank’s compensation framework under the Third Amendment Directions of 2026 is capped at ₹25,000 per complaint, covers losses only up to ₹50,000 gross, and expires on December 31, 2027. From that date, there is no compensation mechanism for authorised push-payment fraud victims.

The MDR revenue—estimated at ₹244.52 billion annually—is directed entirely towards operational sustainability and ecosystem investment. None of it funds victim compensation. P2P transfers, where push-payment fraud is most concentrated and individual losses are most catastrophic, remain outside this pricing framework—unpriced and uncompensated.

The Scale of Change
Annual MDR revenue moves from approximately ₹8.31 billion in 2019-2020—when 0.30% applied to a P2M value base of roughly ₹2.77 trillion—to approximately ₹244.52 billion today: a 29-fold increase driven by a 22-fold expansion of the taxable base and a 33% increase in the rate.

The MDR-liable merchant universe grows from 5 million large enterprises to an estimated 15-20 million—the balance of the 65 million total after P2PM exclusions: three to four times wider, at a higher rate, on a base that was not the subject of any proposal in the public debate as recently as six weeks ago.

The average MDR per chargeable transaction rises from approximately ₹7.10 to ₹33.45, and the ratio of MDR to the unit processing cost of 70 paise deteriorates from roughly 10 times to 48 times.

A fixed two-sided fee structure—a small flat monthly account-level levy on active users combined with a flat bucket fee on the acquiring side by transaction band—would have generated comparable revenue without value-based extraction, left 92% of P2M transactions untouched, eliminated gaming at the ₹2,000 and ₹75,000 thresholds, and created a direct financial stake for banks in fraud prevention through a compensation pool.

Unlike an ad valorem MDR, the account-level levy prices access to the rail rather than penalising transaction size. It would have been more defensible, harder to circumvent, and more consistent with the economics of an account-to-account system.

That design was available. A better way existed. The one chosen prioritises familiarity over fitness, and operational sustainability over the protection of those who bear UPI’s deepest risk.

This is the concluding part of a two-part series. Part I examined why NPCI’s MDR is the wrong instrument, why the merchant base is wider than intended, and why the pass-through prohibition is weaker than it appears.