MDR Leakage Flag Checker: Eight Named Patterns, Per-Flag Exposure
Eight named MDR leakage patterns from the June 2026 India payments research dataset. Enter the volume and rate inputs for each pattern that applies to your settlement profile. The checker fires per flag, computes the monthly rupee exposure, ranks the flags by financial impact, and surfaces the detection methodology and dispute path for the top three. Use alongside the MDR Effective-Rate Calculator for full per-network reconciliation.
Zero-MDR mis-billing on UPI / RuPay debit
Network MDR was mandated zero on bank-account UPI and RuPay debit from 1 January 2020 under PSS Act Sec 10A and Income-tax Act Sec 269SU. Any non-zero MDR component on these rails is leakage by definition.
Premium-card misrouting to 3% slab
Fires when auto-routing applied without a merchant-confirmed BIN to card-product map. Non-premium card volume gets billed at the 3% premium slab and consumer rates do not recover.
Domestic BIN charged at international rate
BIN issuer-country resolves to India but the row is billed at 2.69 to 3.5% international plus forex. Caused by acquirer BIN mis-classification.
Commercial card billed at consumer rate (or vice versa)
A blended slab applied without per-tier verification. The fired exposure is the misclassified volume multiplied by the rate gap between the contracted and billed slabs.
Amex / Diners hidden in blended rate
Fires when Amex plus Diners volume exceeds 1% of card GMV AND blended rate is below 2.2% AND no separate Amex line is present. Exposure equals Amex plus Diners volume multiplied by the gap between 3% and the blended rate billed.
Flat-rate concealing per-network cost
Enter method-mix percentages and the flat rate billed. The checker builds a mix-weighted expected cost using the RBI and NPCI rate matrix and surfaces the gap. Mix percentages should sum to 100; the checker normalises if you under-fill.
MDR not reversed on refunds / chargebacks
MDR is non-refundable industry-wide. On refunds the merchant loses original MDR with no offset; chargebacks add dispute fees. Subscription cancellations compound this every cycle.
Recurring add-on + eNACH rejection stacking
Subscription add-ons (e.g. Razorpay 0.99% per transaction) stack on base MDR. eNACH mandate-rejection fees accumulate on retries at approximately Rs 15 plus GST per failed debit.
Ranked exposure
Flags sorted by computed leakage per month. The top three rows below the table get a detection methodology and dispute path.
| Rank | Flag | Status | Leakage / month |
|---|
Top three recommendations
The eight named MDR leakage patterns
The first three patterns are network MDR mis-billing patterns. Zero-MDR mis-billing on UPI bank account or RuPay debit is the simplest: any non-zero network MDR component on these rails is leakage by definition, because the network MDR has been mandated zero since 1 January 2020 under the Payment and Settlement Systems Act Section 10A and the Income-tax Act Section 269SU. The detection signature is a non-zero charge on a UPI bank or RuPay debit line that is not separately marked as a contracted platform fee. Premium-card misrouting to the 3 percent slab happens when the gateway auto-routes premium, signature, or infinite cards to the premium slab without a merchant-confirmed BIN to card-product map; non-premium card volume drifts into the 3 percent bucket and the consumer slab is over-billed. The detection signature is a contracted consumer card rate of 1.4 to 1.9 percent against an effective rate at or above 2.5 percent on cards that are not actually premium. Domestic BIN charged at the international rate happens when the acquirer mis-classifies the BIN issuer country; an India-issued card is billed at 2.69 to 3.5 percent international plus forex. The detection signature is a BIN issuer-country lookup that resolves to India on a row billed at the international rate.
The middle pair of patterns concentrate on tier misclassification and blended-rate concealment. Commercial card billed at consumer rate (or vice versa) fires when a single blended slab is applied to a mix that contains commercial or corporate cards without per-tier verification; commercial cards carry interchange of approximately 2.5 to 3 percent, and a blended consumer slab cannot recover that on non-trivial commercial volume. The detection signature is a card mix with above 10 percent commercial share against a blended slab below 2.2 percent. Amex and Diners hidden in the blended rate fires when Amex plus Diners volume exceeds 1 percent of card GMV, the blended rate sits below 2.2 percent, and no separate Amex or Diners line is present on the settlement file. Amex and Diners cost 2.95 to 3.5 percent across every gateway studied, so a sub-2.2 percent blend cannot recover that cost on non-trivial volume; either the gateway is cross-subsidising or is reclassifying through a hidden premium line. Flat-rate concealment fires when the method-mix-weighted expected cost diverges from the flat rate billed by more than 0.15 percentage points; a UPI-heavy subscription merchant on a flat 2 percent contract overpays relative to the true mix-weighted cost as the mix shifts.
The final pair are recurring-economics patterns that compound. MDR not reversed on refunds and chargebacks is the most under-audited leakage class: MDR is non-refundable industry-wide, so on a refunded order the merchant loses the original MDR with no offset, and chargebacks add dispute fees (typically Rs 200 to Rs 750) on top of the MDR loss. For subscription businesses with cancellations, this compounds every billing cycle. The detection signature is a refund rate above 5 percent of GMV against a contract that has no refund-MDR reversal clause. Recurring add-on and eNACH mandate-rejection fees stacking fires when subscription add-ons (e.g. the published Razorpay 0.99 percent recurring fee) stack on base MDR without contract basis, and when eNACH mandate-rejection fees of approximately Rs 15 plus GST per failed debit accumulate on retries. For high-frequency subscription rails the rejection-fee tail can rival the base MDR. UPI AutoPay, where the bank-network MDR is zero, is the cheapest recurring rail; steering mix to UPI AutoPay is the structural fix.
The escalation rule for a per-network audit is direct: trigger when the effective rate exceeds the contracted rate by more than 0.15 percentage points across a settled month, or when any single flag exposure exceeds Rs 15,000 per month, or when refund-MDR retention exceeds 5 percent of GMV without a contract clause. At Rs 1 crore monthly GMV, 0.15 percentage points is approximately Rs 1.8 lakh per year of leakage, large enough to fund the analyst time required to run a per-transaction sample audit of 100 to 200 transactions per network. Hold the audit for two consecutive months to confirm structural leakage versus a settlement quirk. Re-baseline whenever the gateway changes its rate card, whenever a promotional pricing window lapses (Cashfree's 1.6 percent flat offer is locked for 12 months and requires UPI at 40 percent or more of monthly GTV; PhonePe's Free promo is time-bound), or whenever the merchant method mix moves by more than 10 percentage points on any single network.
TransactIG operationalises the eight-flag set continuously across millions of transactions per month. The platform ingests settlement files from every payment aggregator (Razorpay, PayU, Cashfree, PhonePe PG, BillDesk, Pine Labs, Paytm), reconciles per-network fees against the contracted rate card, surfaces variance crossings and refund-MDR retention into a structured exception register with the gateway contact and the contract clause pre-populated for follow-up. The platform does not invent fees that are not present in the settlement file and does not publish merchant pricing on the marketing surface. ISO 27001:2022 certified, AWS Mumbai, implementation 2 to 4 weeks for new merchants.
Related
Merchant Fees Insights
Gateway pricing, network MDR, leakage patterns, contract review.
Payment Gateway Reconciliation
Production reconciliation across Razorpay, PayU, Cashfree, PhonePe.
MDR Effective-Rate Calculator
Per-network effective rate vs contracted baseline with flag overlay.
Revenue Leakage Calculator
Combined annual leakage across TDS, ITC, and platform or gateway fees.
Three-Way Match Exception Cost
Quantify PO, GRN, invoice exception burden in manufacturing AP.
Discuss a per-network audit
If a flag fires above Rs 15,000 / month for two months, talk to us.
Frequently Asked Questions
What is a 'leakage flag' and why are there exactly 8? +
A leakage flag is a deterministic check against a single named pattern in which a payment aggregator can over-charge or mis-bill an Indian merchant relative to the contracted rate card. The number eight is not arbitrary. It is the count of distinct, auditable patterns documented in the India MDR research dataset as of June 2026: zero-MDR mis-billing on UPI bank account or RuPay debit, premium and rewards card misrouting to the 3 percent slab, domestic BIN charged at the international card rate, commercial or corporate card billed at consumer rate (or vice versa), Amex and Diners volume hidden inside a blended rate, flat-rate pricing that conceals per-network cost differences when method mix shifts, MDR retained on refunds and chargebacks without contract basis, and recurring add-on plus eNACH mandate-rejection fees stacking on base MDR. Each pattern maps to a distinct settlement-file column or BIN attribute and produces a per-month rupee exposure when the volume and rate inputs are supplied. The checker fires per pattern, ranks the patterns by exposure, and surfaces the detection methodology and dispute path for the top three.
How do I know if Amex/Diners volume is hidden inside my blended rate? +
Run the per-flag check with three inputs the gateway settlement file will already disclose: total card volume for the month, the Amex plus Diners share of that volume as a percentage, and the blended rate currently being billed. The flag fires when Amex plus Diners volume exceeds 1 percent of GMV AND the blended rate sits below 2.2 percent AND no separate Amex or Diners line is present on the settlement file. The economic logic is that Amex and Diners cost the gateway 2.95 to 3.5 percent on every transaction, so a sub-2.2 percent blended rate cannot recover that cost on a non-trivial Amex or Diners volume; the gateway is either cross-subsidising and will revise the contract within a quarter, or is internally reclassifying the volume through a hidden premium-card line that you are not seeing. The exposure is computed as Amex plus Diners volume multiplied by the gap between the 3 percent reference rate and the blended rate billed. To confirm structurally, request a per-network breakdown of the previous three settlement cycles from the gateway account manager and reconcile against the contracted slab schedule.
What is the smallest leakage worth chasing? +
The operational threshold from the research is 0.15 percentage points of variance from the contracted blended rate, held across one settled month. At Rs 1 crore monthly GMV that is approximately Rs 15,000 per month or Rs 1.8 lakh per year, which is large enough to fund the analyst time required for a per-transaction sample audit of 100 to 200 transactions per network. Below that band, the variance is usually month-on-month settlement noise and chasing it costs more than it recovers. Above that band and held for two consecutive months, the variance is structural and the per-network audit becomes the standard next step. The MDR Leakage Flag Checker computes per-flag exposure in rupees so you can rank: a Rs 4,000 per month Amex carve-out is worth a contract conversation; a Rs 40 per month rounding artefact is not. For subscription merchants with high refund rates or recurring add-ons stacking on base MDR, the per-flag exposure on flags seven and eight can dwarf flag five (rate variance) and is often the highest-priority recovery.
Do I need a BIN-tier mapping to use this checker? +
No. The eight flags are designed to fire on aggregate volume and rate inputs that any merchant can pull from a monthly settlement report without a BIN to card-product mapping. For flags two (premium card misrouting), three (domestic BIN at international rate), and four (commercial card mismatch), the inputs are estimated shares — the percentage of premium cards in your portfolio, the percentage of domestic BINs inside the international-rate bucket, the commercial card share — which are usually derivable from the gateway dashboard or a sample audit of two to three settlement files. Refining the exposure number to a per-transaction certainty does require a BIN to card-product mapping against the acquirer's interchange schedule, which is a Stage 2 follow-up once the flag fires. The checker is a Stage 1 triage tool: surface which of the eight patterns are economically meaningful in your settlement profile, prioritise them by exposure, and decide whether the variance warrants the deeper per-transaction audit.
How does this differ from the MDR Effective-Rate Calculator? +
The MDR Effective-Rate Calculator is volume-by-network in, single effective rate out, with the eight flags surfaced as a side-effect of the per-network reconciliation. The MDR Leakage Flag Checker inverts the workflow: the flags are the primary unit, each pattern has its own input panel and exposure number, and the output is a ranked register of how much each named leakage class is costing per month. Use the Effective-Rate Calculator when you have clean per-network volumes and want to know whether the headline rate matches the underlying cost. Use the Leakage Flag Checker when you suspect one or two specific patterns (Amex hidden in blend, refund non-reversal, recurring add-on stacking) and need a per-pattern exposure to size the contract conversation. Both share the same eight-pattern taxonomy, the same expected-rate references from the RBI 2017 circular and the NPCI zero-MDR mandate, and the same 0.15 percentage point audit-trigger threshold. They are companion tools, not alternatives.
Move from one-month checker to monthly named-flag audit
TransactIG runs the eight-flag check continuously against Razorpay, PayU, Cashfree, PhonePe PG, BillDesk, Pine Labs, and Paytm settlement files. ISO 27001:2022, AWS Mumbai, implementation 2 to 4 weeks.