Update/Correction: An earlier version of this piece assumed that the MDR on a UPI transaction accrues entirely to the consumer-facing app that initiates it, especially PhonePe. That assumption was wrong, and it inflated my estimates of what PhonePe and Google Pay stand to gain. I’ve received several well meaning comments on this from folks in the industry I trust, ranging from “grossly wrong” to “needs correcting”. So I’m correcting this piece.
Cui bono?
The Indian government appears to be preparing to let banks and payment apps charge merchants for UPI transactions again, by amending the Payment and Settlement Systems Act to undo the zero-MDR regime it had initiated in 2019.
MDR or merchant discount rate refers to the amount that merchants have to pay to payment service providers for enabling the transaction. It typically gets levied on credit cards, and at one point it time, it used to exist for UPI. Think of MDR as a UPI tax that merchants have to pay to the payments ecosystem.
Following that 2019 budget speech, MDR for UPI and RuPay debit cards has been zero since January 2020, when Section 10A of the Payment and Settlement Systems Act was amended. Now, the government is looking to amend the same section, in order to enable charging of MDR to merchants, with the MDR determined by government notification.
Who benefits?
MDR is not paid out to the app you scan a QR code with. It gets split between every party that touches the transaction: the buyer’s bank (the issuing bank) and the merchant’s bank (the acquiring bank). Then there are the banks in the middle: UPI apps can’t connect to UPI’s rails on their own, so PhonePe operates through a partner bank, and so does the merchant’s app or payment gateway. That’s two more banks in the chain. Then you have the apps and the payment gateways themselves.
A payments industry executive I spoke to expects the pecking order, largest share first, to run as follows: the buyer’s bank, then the merchant’s bank, then the buyer’s app’s partner bank, then the merchant-side partner bank, and only then the UPI app and the merchant’s payment provider.
He was careful to flag that “the sizing bit is just conjecture… but there’s no question that all mouths have to be fed.” Two other things:
First, the apps’ economics are different from the banks’. As the executive put it, PhonePe and Google Pay “make switching fees. Their role and risk is not proportional to the Rupees moved, only the number of transactions moved. The issuer/acquirers’ (banks) roles and risks are a percentage of Rupees moved.”
So because they also make switching fees, both value and volume matter for UPI apps.
Second, he says that the apps already earn a revenue share today, even under zero MDR, so MDR is an increment on an existing line, not a jump from zero. In his words: “the PhonePes of the world already get a revenue share, even pre-MDR… so I don’t imagine they’ll get a much bigger chunk of the pie. The biggest beneficiary will just be the bank that gets the most inward cash, which is likely HDFC, then SBI, then ICICI.”
A few other things:
First, the merchant-side math doesn’t change with any of the above, because merchants pay the full MDR regardless of how it’s carved up within the payments industry. NPCI doesn’t show the P2P vs P2M split on its website anymore, neither does it give the split for each UPI app, and I’m relying on this government press release with 2025 data. The calculations show that:
- Assuming average MDR ranging from 0.02% to 1.5%(unlikely), and all transactions accruing MDR, merchants will have to pay anything from ~Rs. 2,038 crores to ~Rs 152,000 crores a year to payments companies.
2. Of that pool, transactions initiated via PhonePe would generate anything from ~Rs. 1,000 crore to ~Rs 75,000 crores a year, and Google Pay’s from ~Rs. 683 crores to ~Rs 51,000 crores. (Note: The correction to the previous version of this post is that the apps’ actual cut will be a fraction of these numbers.)
3. The single biggest direct beneficiaries, then, are banks, particularly the large banks that see the most inward merchant flows: most likely HDFC, SBI, ICICI.
So was I entirely wrong about PhonePe benefiting? I don’t think so, for three reasons:
1. Even a minority slice of a pool this size is real money for a company heading into an IPO, and it’s a slice of a growing, government-notified rate with future upside built in. It is still a monetisation story to put in front of institutional investors now. For an IPO, the existence and trajectory of the revenue line matters as much as its current size. It also bumps up the valuation. A higher valuation is useful: it grants the ability to take on more debt, and use its equity for acquisitions, which leads to consolidation in a market that has only two players of consequence. That gift from the government of India to PhonePe, and to Walmart, which retains majority ownership post IPO.
2. While UPI apps will not make as much as banks, the two most dominant ones (83%) — PhonePe and Google Pay — are in a position to negotiate better revenue share from banks.
It remains likely that MDR ends up entrenching the duopoly of PhonePe and Google Pay. Of course, it’s not their fault that they benefit from this change.
Why did the Indian government remove MDR?
The government of India potentially stands to lose here. The benefit of zero MDR to the state was never really about consumer convenience. Because UPI is treated by users and merchants as equivalent to cash, merchants were effectively required to accede to the customer’s demand to scan a code rather than hand over cash (notes).
The growth of UPI therefore digitised merchant receipts, and digitised receipts create a money trail: it now lands in a bank account. That gives the GST department proof that they can use to send tax notices to small merchants. Tax related surveillance is what the government actually gains from a cashless economy.
The reintroduction of MDR gives those merchants further incentive to try and switch to cash, and in case they decide to ask customers to pay the MDR, eventually gives customers the incentive to switch to cash.
So the shift tells us two things:
First, that the government of India has given in to sustained lobbying pressure from the payments industry, and maybe even the US government.
Second, that it now believes, or has been convinced, that consumer habit is entrenched enough that neither users nor merchants will actually leave UPI, whatever it now costs them.
How much MDR will actually be imposed? How will this impact competition?
Nobody knows, but it’s likely that it will follow the usual arc of taxation… because MDR is nothing but a UPI tax that merchants pay to payments companies.
It will start small. A rupee or two on every thousand, small enough that a merchant barely notices it on a monthly statement, and then it will climb gradually. What it will not be is a purely market-determined number. The RBI may probably set MDR, which ideally should have been the prerogative of a payments regulator, but the RBI has still not created an independent payments regulator, but that’s another story for another day. (Another story for another day is the fact that RBI has still not initiated competition for UPI, which it had planned as “NUE”).
If it were genuinely left to market forces, instead of government notifications, the outcome would be more interesting.
- You would get price competition on MDR, and merchants would begin refusing payments from the apps that charge more, which would be the first time in UPI’s life that one app is not simply interchangeable with the next. Obviously, the government doesn’t want that to happen.
- A small player like Paytm could offer zero MDR to merchants, by treating MDR as a marketing cost (which is the creative accounting tactic that e-commerce companies used when offering discounts), and buy share from PhonePe and Google Pay.
The flip-side is that someone other than big tech firms like Google, Walmart, Meta and Amazon now has a business case for a UPI app. There’s actual money to be made in UPI, and there’s a reason now to compete to gain marketshare.
For now, regulatory issues (if any) and market conditions (tricky) notwithstanding, PhonePe and Walmart have been handed a great story to sell to institutional investors by the Indian government. As do banks.
