The new UPI equation

Last month, Indians made close to like 24 billion UPI payments. What. do you want to take a guess? I have no idea. Pranav, do you want to take a guess? Eh, last month, about 10,000 crores or so? 10,000 crores, okay fair enough. It's, I think, the number that I have in front of me is 30,000 crores. Did you know it was 30,000 crores? No, that's humongous. That's humongous. And I'm going, my guests are also like a little surprise, but honestly, I was surprised because I just saw it right now. But across FY26, that was 250 billion transactions worth, well, you just have to multiply it by 12, somewhere close to like 314 lakh crores. So, right. Do you want to take a guess of how many people are making these payments in India? Bavik, come on. This, I'm sure you can crack. I'm pretty sure that it's like about 150 million odd. 150 million, enough. I'm going to be close at about, uh, 500 million, 400 to 500 million. Oh, okay. I think, yeah, it's closer to like 400 million because it's actually like 55 crores, 55 crores is closer to like 500 million people. Yeah. And they have been paying close to like 6.5 crore merchants, right? Wow. So, this is like ridiculous. Like the numbers are like quite out there. And there's one very interesting point about every single one of these transactions, which is that until now, every single one of those transactions cost how much? Nothing. They cost it. They cost zero. Yeah. And that was because in January 2020, there was a bill that was passed. And since then, they passed this rule that said that you cannot charge anything for UPI transactions in India. Now, that changed on the afternoon of 6th August, four days ago. We are recording this on, uh, the 11th of August, it was five days ago. And that time, the looks of a pasta bill by voice vote, which basically gets rid of that guarantee. So, we don't know what it's going to cost yet. But because it's, the rules are going to be framed and they're just being framed right now. But the equation that India's payments ran on for like six years has changed. And there is a new one that's going to be written. Today, in this episode of 2x2, I am going to basically have two wonderful guests with me, where we are going to discuss what was this old equation? And how did it really work out for UPI between all of these players? And what exactly is this new equation? To be clear, we don't know what this new equation is. And I have both of my guests with me and we're going to sit together. And we're going to try to figure out all the various possibilities on how UPI is going to change, who gains, who loses, and how this gets split up between them. That's really what today's episode of 2x2 is about. Now, in order to do this, I have two wonderful guests with me. My first guest is Pranav Gundapalai, who is director and senior research analyst at Bernstein. Hi, Pranav. Welcome to 2x2. Hey, Ravi. Thank you for having me on this. Pranav, I want you to tell us the three reports that you authored on behalf of Bernstein. Just read out the headlines for me, please. Because I have seen these reports and that's really why I reached out to you and said, Pranav, you have to come on this episode to talk about this. Right. So I think the first one which we wrote, almost three, four years back, was title as the long view, and said, zero MDR is here to stay. Correct. And then I think we've written a more recent one, which is about a year, year and a half ago, where we said, it's a billion dollar question that remains the most important one in the payment ecosystem. And it will be a month back. We said there is a sense of Dijavu because the MDR on UPI debate is back. But we felt it was closer to reality than ever before. Yeah. I think those were three kind of reports that we had over the last four years now. And I really wanted Pranav on this episode because it was essentially that arc. See that arc. It started off with, oh, whether it is basically here to stay, zero MDR. And now it is well, okay, this time closer. Within like three years, things changed so dramatically in the Fintech ecosystem in India. My second guest, who we have already heard before, is Bhavik Kall. Bhavik Kall is the ex chief broad officer of super money, which is Flipkart's Fintech and the fifth largest UPI app. And before super money, Bhavik spent close to like six years in a range of Fintech companies. And these are companies, he was the CPU of pay you finance where he ran lazy pay and pay sense. And so he's been on the side of payments and monetization in a world where MDR always existed. Now, Bhavik, welcome to Dubai too. Hi. Thank you, Praveen. It's really nice to be on the show today. Yeah. And I want to ask you, Bhavik, because one of the things that you, I'm sure you explain this a lot. So I'm going to ask you to do it. Can you explain MDR to our listeners? Because that's really the context on which everything else that we're going to discuss today comes about. So let's talk about MDR. If you have to explain MDR, what is MDR here? Yeah. So in basic language, it is the fees that a shopkeeper or any business pays to enable a transaction to go through. So assume that transaction is a product or service. And the shopkeeper or the business owner needs to sell a product or service. So he's paying for that service. And that's what MDR is. It's the fees. And usually it's a percentage of the transaction. Yes. In majority of the business, it's a percentage of the transaction. That's how it sort of like slopes well with the value of the transaction. In some specific cases like bill payments, it's actually like a fixed value. Got it. So if I asked you to pick, say, a percentage of transactions are, say, for debit cards, credit cards. Could he give us like a broad overview of what those percentages would be? So those percentages run anywhere between half a percent to two percent. Okay. The entire idea here is that the business owner wants very high value of transactions to go through. Which is also called average transaction value or average order value. And once users to come back. So the convenience that the process that enables that to happen is through credit products. And credit products make money through moving the moving the money around right? And so that's the reason they charge more transaction fees or MDR. And that tends to be on the higher side and debit products. Which is money that goes directly out of your bank tend to charge lesser MDR. Got it. So that's the range. Got it. And the MDR for UPI until five days back was zero. Was zero. Okay. And that's really the thing that has changed. So what when you talk about MDR now for the rest of this discussion because you're going to keep talking about MDR again and again. Yes. This is really what you're talking about. So I just want to very quickly like talk about this. Which is since January 2020. Like I said on UPI and on RUPI debit it was zero by law. I think to kick this off, let's start with what I call the old equation. And by this old equation for both Prana and Babik runs from like 2020 to 2026. Prana maybe you can start off because I think you have done some a lot of like very interesting tables that I read in some of your reports on this. Can you walk us through what was the old equation? Like in a world where you basically say to merchants and to everybody else that the thing is zero MDR. How does this whole thing split up? What was really happening with respect to fees in the old world? Okay. So the fees part is easy it was zero right so nobody paid any fees at least for the merchant transactions of on UPI. Now this again was a big shift away from card based transactions where there was always a fee involved. So this was literally a new regime or in that six years ago that was a new formula. Now at that point it was a bold experiment from the government and I think there were a lot of factors which allowed this to happen. I mean A because they just passed a law and two also because RBI could get the banks to agree to this. You've seen this play out in many countries where such an experiment failed because the banks simply said hey we lose our debit card fees and therefore we'll never do this. So banks actually got along and did this as well. And I think having a large PSU banks helps because they just do what's right or what's read it by the government as well. So it all happened I think as you know it started it was still a very very small percentage of transactions when it started. So he said yeah I can put a zero cost transaction it's still negligible percentage of my total transaction so why not. But then you saw an exponential growth I mean growth that we have not seen in almost anywhere else in the world where we went from you know. Being a lagard in terms of digital transactions to one of the leaders and digital transactions. Now as these started to scale you know you started hearing complaints about how there was so many people in what incurring costs for these transactions which is not getting compensated and some of those discussions came up. But what was interesting is that the benefits from a shift to cashless transactions actually created. Far more value than the cost it took to provide those transactions. give you a few examples. Yes, please. So today, you know, RBI's own estimates is that the cost per UPI transactions about two rupees. So that's, that's the cost. And therefore, when the government pays that 2000 crores or so every year, there is a complaint that who that's, you know, hardly 10% of my of our industry costs, whereas we are incurring, like I said, let's add 20,000 crores to facilitate UPI. Got it. So let's stick with that 20,000 crores. So that's what it's costing. Hold hold hold, once again, I want you to break that down. So let's just go one by one. Okay, so now we are basically saying, I think let's talk about this stakeholders for a second. So the first stakeholder is of course banks, right? Right. The second is of course the payment apps, which is places where Bavik has worked at. The other one I assume is consumers themselves. That's a third stakeholder. Yeah. And who else is there? Service is a government. Government. There are several service providers in the middle of the guys who design software at NPCI, the switches, people who enable transactions to go through. Okay. Got it. Let's take these four five for a second. Okay. Now let's start with banks. Now, as you said, but enough banks, if you're just talking about the old before 2026, now there is certain things that bank, let's talk about the cost benefit for the bank site. Now here are stuff that I can imagine of the top of my head. Number one, as you correctly said, banks lost out on a certain amount of money because whatever UPI transactions is sort of enabled cannibalize their debit card and credit card transactions. So they lost some money there, right? That's one. Right. But what else is there in that equation for banks specifically? Yeah. So debit cards had just start to become so we were never really seen, never really seen a big scale up in debit cards because it is expensive and card infrastructure, accepting infrastructure was weak. So I think that's a small revenue pool of that lost. But what they saved is actually massive. So let me put some numbers there. So in India, we were doing about 10 billion ATM transactions in a year in 2019. And today that number is dropped to five billion, right? So in a six or seven years period, the number of accounts has grown, the population has grown, yet the total number of ATM transactions has come down by 50%. And each ATM transaction costs about 20 rupees give or take. So which means banks, even if that same, you know, 10 billion had stuck, you would have saved almost about 10,000 crores simply on reduction in ATM transactions, right? And if it had grown in line with the growth in, you know, in the number of accounts or number of consumers, then that could have easily almost doubled from where it was in 2019. So there is a savings there. Now it was not cost that they were incurring earlier, some part of that is gone, but they didn't have to spend on an expanding ATM network. So the ATM network has been flat for the last seven, eight years. So that's a massive savings. Interesting. I never saw it this way because I think this is the single thing that changes. Like even before this, when you spoke about international examples, you said that, oh, banks would never really agree to a new kind of a product which has like zero fees or zero MDR in this case, because it fundamentally cannibalizes an existing revenue pool, which is either debit cards, so credit card, like how Bavik described, which is like a certain percentage of 1.52 percent, et cetera. But in this case, you're saying because the number one, the number was so small, but the ATM said, so net net banks saved money as a result of UPI? Paradoxically? Yes. They actually have big beneficiaries. Probably there is another part. What is the other part? So one thing that banks and you're right, prana, may not have like foreseen at that point is the scale at which transactions on UPI might have run. And the infra requirements on the technology side would that would have been required to like run technologies at that scale. So the cost that has come only on the back of running these, you know, transactions worth 30,000 crores per month has been quite humongous on the side of the banks. Just to build it out, scale it up, the existing infra, the core banking systems at the banks were not actually ever designed to like bear such a heavy load. So that is the one cost element that nobody had thought about, and which is also take the reason why UPI for a very long period of time had seemed really bad performance, right? Every every third transaction used to go into like a progress state, every time they used to be disconnections. And now you might have seen after like a long time that the success percentages are like almost as high as 99, 99.5 percent, right? And that's come on the back of a lot of investment that's gone on from the bank, at the NPCI and at the apps, all three years. Interesting. Yeah, you're right. I don't two years back during IPL. They would definitely be a time where that is when the success rates would like really plowed, but you're right. Now, off-late, I have not seen too much. Yeah, so I have a very interesting anecdote. This was when IPL was just, you know, starting to gain traction about five years back, and I was talking to the guys at dream 11. And interestingly, the tech that they had built to ensure that routing happens was state of the art. Because at scale, banks were supporting probably about 200 or 300 transactions per second. And these guys were processing even in those early days, something like 7000 transactions per second. And there was no. So as soon as IPL matches used to start, anybody running an IP API transaction would see downtime. And that required a completely different set of engineering infrastructure to be worked on. Interesting. Okay. Okay, Pranav, do you have any, either if you have any number, I know it's not easy to estimate this, but I'm just trying to take a guess. Could you estimate this with the amount of, you know, the investment made from the asset for the technology of UPI from the bank side? Pranav, is it a significant number? See, there are two parts to it. So one is, there is in us, there is an RBI paper, which actually puts out the cost of UPI transaction. And if you sum it up, then it works back to that 20,000 crores or so that the entire industry spends, which is a combination of service providers, payment apps, banks and all of that. Then there is also the part where I think banks were playing a bit of a catch up on tech investments as well. Because I think this coincided with a shift away from physical branches to digital transactions and digital interfaces. So it's, I think at a bank stage, it's hard to segregate how much as it was specifically for UPI and how much is it simply because things have scaled up and, you know, you had the whole jump done program in the past where you added accounts, now they started transacting, suddenly the numbers start going up. So some of it was organic growth itself and some of it was because a number of transactions per account also went up. So let's assume, let's take the bank side of the equation. But if you remove the technology part or at least you keep the technology thing on the side, the investment on the side, what you're saying is there is a certain amount of savings that banks got paradoxically because people use the ATM's less and ATM charges went down, etc. Let's take Bavik's point into account where they say, okay, there is some technological investment they did. Let's assume that at the end of the day, they are probably, I think I would like to believe they're a little bit in the red. I wouldn't say that they're significantly in the red, but let's assume that they're a little bit in the red to the extent that you can attribute it. That's one. Now let's talk about the second stakeholder, which is the government itself. Now, this is where it comes down to a subsidy. And now, Pranav, you should probably talk to us about this subsidy. Okay, see, the government has, I mean, today pays about 2000 crores as subsidy, which seems like a generous thing to do. But actually, they have saved a lot more on currency printing. So if you look at just the currency printing costs, forget about improved taxation and all of that, the currency printing cost used to pretty much move in line with GDP because your cash required was almost in sync. So back in 2018, 2019, they used to spend about 50 billion rupees in printing currency. That number without UPI should today be closer to 100 or 150 billion, but surprising or not so surprisingly, that number is actually declined marginally over the last seven years, which means at the minimum, the government is saving about five to 8,000 crore rupees on printing currencies. So 2000 crores of subsidy given, but have already captured 7,000 crores of cost savings from lower currency printing costs. And on top of that, I also get better tax compliance because it's harder to, it's easier, it becomes more transparent, etc. So tax efficiency goes up. So it's a great win for the government. Interesting. So for the government standpoint, big tick, green tick, right? You're basically saying, okay, it looks like they're down 2000 crores, but they're not. Like I said, there are these other factors and I don't think, is there anything else from a government standpoint that you think we're missing? No, make sense. Fine. So the government is like a big tick. We'll come to this subsidy, where does this go to? We'll come to that. But now let's go to the third stakeholder. Third stakeholder is your favorite, Babik, which is the platforms themselves. Okay, talk about the platforms. Yeah, so it's on the platforms side and I have done some homework on the economics. It's very interesting. So just so that your listeners can understand it better, I normalized it to 100 rupees spent. Okay. So at 100 rupees spent, every 70 rupees is towards P to P, not P to M, which is very. Explain the intuitive. Explain P to P and P to M. Okay. P to M is person to merchant. That means a person paying a merchant and this merchant can be a school Kiranah store. It can be your favorite deed. to see brand online, and P2P is person to person. This is me paying my friend, my mother, my wife. So interestingly, if you look at volume of transactions, P2M transactions are much larger in the ecosystem. So by a factor of, and let's just take it roughly, about 70, 30. So 70 are paid to merchant and 30 are paid to people, right? But when it comes to value, it's the opposite. So the people to people payments are much larger in value than the people to merchants. So the seven out of the hundred rupees spent across the ecosystem at any point, out of this 30,000 crores per month that you were talking about. 70 rupees are sent between people, and 30 rupees are sent from people to merchants. So that's interesting. Now, if you were to, and maybe Prana will talk about the finance ministry call out on, what are the rules about the transaction value, the size of the merchant, et cetera, et cetera. But if you look at the transactions which make to this MDR cut, that would be lesser than half. So out of this 30, about only 15, generously would be the ones who would earn the MDR. - So once again, when you talk about MDR, you're basically saying this is below the 2000 rupee mark. - This is about the 2000 rupee. - About the 2000 rupee, fair enough. But we'll come to the 2000 rupee. Forget the new regime for a second. Let's just try to understand what the economics was earlier. So now in this world, if you were, say, one of these apps, let's say you were the Google Pays or the phone Pays or the supermanies or the Navis of the world, what is happening to you in terms of your ledger? - So let me just break down the unit economics. So on the cost side, interestingly, it's not as different. So on a 15 rupee qualifying spend, you're both spending between one to two Pays. This is normalized. - Got it. - And take it with a pinch of salt, not exact. On the revenue side, there is an interesting case. The apps make slightly more revenue because they are the originators of transaction. So they get to negotiate heavily with banks, but not as much. So it's 0.5 Pays on the bank side and one Pays on the app side. But the biggest cost difference that comes into play, which is what we'll get into later, is the fact that apps have to acquire, retain and engage the customers. Now for a two Pays cost, the apps are spending almost 13 to 15 Pays. And so the acquisition out of that is about two Pays, but the retention, the rewards, the re-engagement, which is what P or all sort of like you used to write, like that two rupees that you get on a 500 rupee spend, because that is where this business is so heavily skewed to the incumbents. And that's the reason we'll talk about later, how challenges find it so difficult to get into the space. So even after the MDR, if the MDR world to come in and you will talk about the range, right? It actually is fairly negative on the app side, compared to the banks. The banks is where this will start to make much more sense than it will for the apps. - Where are we getting ahead of us? I don't want to get to the MDR and how things are going to change. I want to stick to what was happening. So in the old world, just without that, so your app will end at some negative, say, minus 14 Pays for transaction, and your bank would have ended at some negative 0.5 Pays for transaction. - Got it. Pranav, walk us through the old world with respect to these apps, which I'm assuming are the ones who are really in the red, but let's talk about it. - Yeah, so I think they have a right to complain because there is no direct monetization from these transactions, right? So it's a bit like also similar to the banks where from the transactions, yeah, I don't make money. There's a small fee that they probably earn on the P2P side, but nothing much, nothing material. - That's really what they call interchange, right? It's essentially like from my network to your network, if I pass it on, you give me a little, but from your network to my network, when I come, I pass it on a bit. I mean, I would like to believe they're fairly negligible in the larger things. - Negligible in the larger scheme of things. I mean, the volume might make it slightly, but yeah, the broader scheme of things, that's not what you're going to build a business around. - Sure. I think what has happened is a lot of these platforms, I mean, when they came in, they knew that this was a zero margin business. So they came in knowing that this is a zero margin. And I think what the driving factor was, they've seen how some of the models have got built in other markets, where you can use this to cross sell a lot of things. And all these platforms have tried doing it. So the entire payment space, like what I've been mentioning, is it's more than just paying for attention. The payments itself was a hook to get consumers on board. And then, I think the platforms tried selling them a lot of things, mutual funds and credit and whatnot. And on the consumer side, they've had limited success in cross selling. On the merchant side, though, I think they've had a much better range of outcomes. So they have managed to sell them these sound boxes of spot speakers. They've managed to sell them loans. So that is where they're trying to monetize. So even though this remains a zero MDR regime, a lot of these platforms are close to break even, or have already break and even. Which tells you that, yeah, I mean, this is a cost, but there is money to be made elsewhere. It's a bit like, bang, saying, hey, I'm paying you interest on your deposits and I'm paying you giving you free ATM. So I'm kind of losing money, but you're using that to sell a lot of stuff to the customer. And hence you could say, yeah, you should look at it in aggregate fashion, where they're kind of figured out a business model. So not so bad for them, but yeah, specifically for payments. Yeah, I mean, they're not making money. I mean, Bavik is going to vociferously disagree. Okay, I see it. No, but I'm saying, but Bavik point taken, which is that we agree that on the payment side, your numbers seem correct, which is that you're losing, you said what, 14 by say, give or take. Yeah, I'm saying. On the apps and the bank does not lose as much. We'll come to the banks, the banks are done. So we'll talk about the apps. So the apps are essentially losing close to like 14 by say per transaction. Yeah, but what about this point that Pranam makes, which is that even in this world, where MDR is zero, all of these apps were able to make a certain amount of revenue either through cross-selling, whether it is through loans or insurance or other Fintech services, either on the merchant side or on the personal side. Do you feel like that broadly fits? So I think I don't completely agree. Number one, were they able to make revenue? The answer is yes. You can see the likes of ATM doing seven and a half thousand cross, phone pay doing eight and a half thousand cross. So yes, they have been able to make revenue. Have they been able to make revenue profitably on the back of this product is still debatable. So if you were to look at just these two businesses, because one is in the public market and the other is just going to get to public market, so there's enough details about how they function, is the difference between one being cash flow positive and profitable, which is I think ATM just declared about 500 crores of profit, versus the other being as much or slightly more in negative in losses. And the difference is the percentage of lending that contributes to this revenue mix. In ATM's case, it's much larger. For phone pay, it's about 90 odd percent comes from payment services. And for ATM, lending contributes about 30 percent. And that's the reason that one company is doing well. But while you put it out there, the lending contribution, like Pranov mentioned, is far more on the merchant side than it is on the customer side. And that's the reason that ATM has been able to turn around this ship. The cross sell percentages across the board for some of users at Bharat scale, like you're talking about, say, 400 million users coming into the ecosystem. And phone pay has like what, about like 150 million users on a daily basis, transacting at that scale, getting them to actually use a credit product is really hard because there isn't enough information about their previous lending history, their transaction history, et cetera, out there. A lot of this users have come into the banking system quite recently. So the cross sell rates are fairly low because everybody is trying to sell the same people, the same 30, 40 million or 100 million people, the same credit products all the time, right? So on the consumer side, while there has been some revenue on the lending piece, the larger piece of the pie is definitely on the merchant side. Got it. So what are you really saying is if you take the payment app side, and I know that you vociferously defended the payment apps or actually gave more context about the payments, which I understand. No, I think the argument is that it's kind of like a depends. I mean, that's really what you're saying. We said that government is a big tick. We said that for banks, it is essentially like somewhere in the middle. It's a big tick if you remove the technology cost. If you add the technology, or maybe they're a little bit in the red, for payment apps, what you're really saying is it depends. And the reason why you're saying it depends is because it depends on two things, which is that, I mean, premise point is that, oh, you can take it and you still make money. You're like, yes, but it depends on where you make money and how you make money. Like PTM, as we have established, is much stronger on the merchant side. Because it's stronger on the merchant side, it's able to make money better there. Phone pay is most stronger on the consumer side, and the jury is out about whether it can make money through these other additional services on the consumer side. So it depends. Okay, fair enough. The fourth one is easy, and I know that the two of you will agree on this. And the fourth one is consumers. Which is pretty straightforward, right? I think it was zero. They pay nothing. Have they ever paid anything? Never, right? So it's zero completely. Any other stakeholder? I think you also mentioned NPCI and the others. Do you want to talk about the, is there a cost-benefit equation for them earlier? So it's very interesting, NPCI created the center protocol enabled banks to create like some shared sense of like knowledge repository of what do mobile numbers linked to accounts mean what do these virtual private addresses mean how do they link to each other, how do banks talk to each other, there wasn't any shared communication channel and that for that reason they do make some money but over the period of these years that that that volume has really grown. So, because you can imagine that many number of transactions go through the entire information rail, also NPCI is technically a nonprofit, I think they they make close to 1000 500 cross in annually not as revenue revenue is slightly higher than that, just in like what would you call profits but obviously nonprofits can't recognize this as profits. But that's the scale so NPCI has been in the big green take in my opinion. No, I think you need to see it in two to three different lengths, I think one is for the scale of transactions that they operate for the network that they provide, you could make the same argument for any of the global card networks visa or master, they pretty much do the same thing, but the actual amount of surplus that they make is miniscule compared to what the global payment networks make. Yes, but there were the ones who were willing to build the infras, so I guess it's fair that they make that surplus and second, I think what is help NPCI is that they have even when they started off on the UPI journey, I think they already had other properties, beat the ATM switch or beat the Rupin network beat B to beat transaction platforms, etc. So this became a bit like that venture investment that you did, which then made pretty stellar returns for you, so wouldn't really see that as excess for what NPCI did. Okay, interesting, I have a point though, now let's just now let's just finish this, we are basically sort of gone through the old world and we've gone through every one of the stakeholders and we have understood this so just to recap, right, number one banks, okay, slightly in the red, government, a big green platforms, well, when you both are disagreeing, but let's just say, okay, fine, it depends, consumers and merchants, big green, because it's zero, NPCI definitely, big green, okay, great. Now it seems like broadly things were working, right, now explain to me this, now I remember that at a certain point in time, one of the things that happened is there was this figure of 20,000 crores. Yes, but now I want you to explain this 20,000 crores, so let me just broadly say what it was, it was that somewhere this figure was put out that said that, you know the cost that we incur to do this entire thing to keep UPI free for all of India is 20,000 crores, which is more or less the pressure that they put on the government in order to do this. So let's talk about it, who is this, what is this 20,000 crores, who is spending this 20,000 crores, how was it calculated, do we know anything about it? So I think the origin for that is there was a payments white paper that RBI put out, this was back in 24, if I remember right, where they, you know, it was a classic white paper where it didn't take sides, it presented both sides of the argument. And part of that discussion in as part of the discussion, they put out what it costs to enable UPI transaction, I think the number was about two rupees per transaction, and that extrapolated into the number of transactions is probably what takes you to that 20,000 crores, etc. Now, I think the part of the argument back when RBI wrote this is to say the cost is almost one third or even one sixth of what a typical debit card transaction costs, and therefore it's, it's okay, I mean, it's not as bad as giving out card for free, a card transactions for free. But then that got extrapolated into saying, hey, this is what it costs, nobody is paying for it. And I think where it started, I think becoming a bit more vocal is when you had many of these, you know, almost 80% of the cash transactions going through UPI, it was not just the small merchants, but even the large merchants who were earlier using, let's say, card transactions, migrated to UPI, and there they saw it as a cannibalization. I think initially, it was sold as, hey, cards will always be there, this is, you know, for really small ticket transactions, small merchants, let's make it free, it sounded like a nice thing to do. But as it scaled, the really large merchants also started seeing bulk of the transactions comes through UPI, and then these platforms and banks were like, hey, that doesn't seem fair. Why should we bear that cost? Let's put a number to that cost, it's 20,000 crores, so that's where it came from. By the way, I must say that that number keeps changing, okay, like I think the other thing that I know is that in March 2025, the payment council of India sent a letter to the Prime Minister. And in that letter, they basically wrote that running UPI cost 10,000 crores here, now it's 10,000 crores, okay, and that letter, they asked for a 0.3% MDR on large merchants, and this was in March 2025. So I think their argument was that look, you're giving us a subsidy, but that subsidy is only covering like 10 to 15% of the costs, right? Do you have a perspective on the subsidy and the cost that is being incurred by the payment apps? Do you feel strongly about it? So interestingly, the value that reaches the apps, there's a fraction of the total subsidy that's put out there in the ecosystem. Yeah, we have to talk about that. What happens to the subsidy? Who gets the subsidy? Yeah, both of you have to tell me about this before we move on to everything. Yeah, I think the Prime knows more about it, but the acquireer banks, the people, the banks who run their transactions, they get the subsidy and then they dispose it across the entire network from there to the apps, the switches, the NPCI everybody. So the banks get all the money. Am I right? Yeah, banks get most of the money, right? Because their argument is, have you are doing all the hard work of doing the KYC for the customer and for the, you know, enabling the payments for the merchants, et cetera, and therefore we get to keep bulk of it. And by the way, this is very similar to a debit card transaction where we keep bulk of the economics, so it should be no different here. Which is also a bit of going back to the way payment instruments have historically worked, right? Because if you see checks or if you see debit cards, credit cards, it's always like the bank which has caught the customer first takes lion share of any revenue coming from the customer. And they pretty much took the same allergy and then extrapolated that just to give you a sense out of the 1500 or 1700 crores, which was like given out in FY25, et cetera, the payment process, the apps, the payment for Infra owners who contributed to almost 90% of transactions, didn't make even one third of that money. Wow. Okay. So I can, I can see banks to, banks to keep the lion chairs. Yes. I can see that they would not be very happy about it. Okay. So then is it fair to say that the people who really wanted this to change were the payment apps? Because if you think about it, we always spoken about banks, right? Even without this subsidy. Yeah. Like Pranava has established that they have, they're definitely like there is a green element there, right? Sure, your point is made about they may have a technology infrastructure fair enough, but look at the savings that they make in ATMs, et cetera. So they're okay. They can live with this. But the people who really were wanted this to change, I would imagine how the payment apps see if you look at banks as a large set of lines of businesses cross lines of business benefit for sure banks can easily justify this, you know, investment. If you were to only look at UPI as a line of business, then of course banks would like market as a red and would like to have a little more money. But for financial processing apps for sure, they are in such red zone. Not any line of revenue, which is added to their PNM makes a lot of sense. And just to be clear, financial processing apps, you mean the Google pays, the phone pays, the navies and the supermanies of the world. Interesting. Okay. So now this is this world. And this was the world until 2026. And now we understood how each of these players had their own incentives and had all of this money broken up between them. Now boom, five days back, this changes, right. And we don't know what's going to happen. Right. Product walk us through what is the number on the table before we will go to each one of them one more time. But I guess we will have to start to speculate now on. Okay. Now this has changed. How should we think about this change? What's the most likely eventuality? See, I think there are two variables at work here. Right. So one is there would be a fee charge the merchant. I think that we will discuss what those levels could be. The second is what type of transactions will come under an MDR. So meaning is it about 2000? Is it only for large merchants? So there is a percentage of value that will get subject to MDR. So those are the two kind of unknowns or variables that need to be determined on the first one. Now obviously there's no indication from anyone like there's a letter that you said 30 pips is what the payment concept of India asked for. There are a few data points that we can use to guess what that range could be. Right. So at the bottom end, you have the subsidy that the government pays by the way. There is some formula to it. Even though it almost seems arbitrary. There is a 20-bips that the government uses to come up with that subsidy value. So that 20-bips in my view becomes a floor. But then the RBI used to provide an additional incentive, which is called the PIDF incentive, over and above this, you know, a 2000 close to 20-bips incentive. So something about 20-bips seems likely, that is the floor. At the upper end, you have a debit card MBR, which is close to 90-bips, right? So that becomes the upper end, right? I like it. So I am between 6 to 12 feet tall. OK, great. All right. OK. It's a very wide range. Now, so that's the absolute extreme. So now let's put more real estate. I think the RBI had put out, again, that payment discussion paper that we talked about, at the cost at about 25 basis points. Somebody been 25 to 30 basis points, is the cost of enabling a UPI transaction. And it might be fair to say, hey, whatever government comes up with, should at least get to 30 bips. So because then you cover my costs. The other way to think is there is a mode of payment, which is UPI, which is 1/3 the cost of a debit card transaction. And debit card is 90. So 1/3 of that puts it, again, at about 30 bips. The other number is for small merchants, debit card costs about 40 bips. So again, that's one more number into the mix. So if I were, if I was a betting man, I would say somewhere between 25 to 40 bips is where it ends. We have assumed it could be at about 35 bips. That's a number that we have thrown in. But somewhere in that-- Fantastic. Fantastic. I mean, you're really going to get your job right off. Because you're obviously triangulated from multiple-- because you're really working from a place of uncertainty. I mean it. Because you really don't know what's going to happen. But you're looking at multiple data points and saying, look, though, if you triangulate from all of these sides, it looks like the number we hit is somewhere close to like 30, 35 bips. Yes. 30, 30, 30. Let's take that and go forward. And I think it's reasonable to assume that barring anything really weird happening it's somewhere in that's the number, OK. But that number is still the-- You're talking about the broad number. Headline number, right? Now, I think I'm going to pull you in barbing. This is, I guess, the part where you bring in the P2M and the P2P, et cetera. If we basically take this assumption, and we say that look, it's going to be around this. How does this break up with respect to the types of transactions that now get subjected to this? So going back to the two numbers that I gave you, that without the MDR, the app came down to say about like a minus 14 passable transaction. And the bank came up to like some minus 0.5 passable transaction. So clearly what now happens is that if you were to assume this is this 15 rupees, and you got to like-- Yeah, let's talk about it in a blended way first. Then you split it up. Yeah, yeah, yeah. So in this 15 rupees, the bank's going to take a-- 15 passable. 15 passable. No, no, 15 rupees of the P2M transaction. Oh, that is eligible. Got it. Out of this, you are now charging say 3,540 bits, right? That comes to like about six passable. So the entire ecosystem per 100 rupee transaction will get six passable. Now how does this six passable get divided between banks and apps is the big question, right? OK, hold. I know you want to fight this out. OK, we'll do this out. But let's talk about the six passable. Now, six passable per 100 rupees of transaction is the additional revenue that is in the mix right now. Let's not talk about who gets it here. But let's talk about that number. Pranav, if you had to estimate the size of this, he's talking about from a per 100 rupee thing. But I think you have done some sizing to say what is the potential profit pool or money pool that is on the table right now. Yeah, see, a couple of things. One, our view is they would keep the P2P out of this because you can't charge consumers. I think charging consumers would be a strict no-no. And if you are not charging consumers, then I don't think there is a big revenue pool on the P2P transactions. So money will come only on the P2M transactions, which today is roughly about 100 trillion. So if all transactions are subject to that 30-bps MDR, then you're talking about a $30,000 crore revenue pool. That is if all transactions are subject to it, right? Correct. Now, I'm sure that's too much to ask. So they would put in some restrictions on size of transaction, size of merchants and all of that. And the good part for the payment apps or the banks, et cetera, is that the transaction values are super skewed in the UP ecosystem, right? So I think a buyer was making a mention earlier. The transactions about 2000 rupees are only 4% in volume, but almost account for 67% of the value. So let's say the regulation comes out and says, oh, don't worry, we're not charging the low value, ticket size transaction, just about 2000. Only four out of 100 transactions will be charged, but you will still capture 67% of the value, which means that 30,000 crore I talked about will become about 20,000 crores, too, very size up. And surprisingly, matches that 20,000 crore that was floating around, or maybe not so surprisingly. Interesting. OK. So looks like 20,000 crores is the number we're playing with. I guess the 15% that you are talking about, Bavik. I guess that is that also like an indication with respect to greater than 2000 rupees? Or is that greater than 2000 rupees? I was greater than 2000 rupees. OK, great. So by the way, I have to ask both of you, why 2000 rupees do you know? Like why is that the number? Why not 1,000 fired? Why not 3,000? Why not 4,000? Do you know? Is there a reason? I'm not even sure whether that's been already different. No, nothing has decided. We are talking that as a reference, but seemingly like it's a reference that has been floating around. So I was wondering if you have a point of your why that reference. We can assume it and move on. No, the only reference is that NPCI, when they disclose their data, used to segregate it as below 2000 and about 2000. And in the recent few months, they've also said 0 to 200, 200 to 500, and 500 to 2000. They brought up a bit more of a granularity. We are really reading the tea leaves here. So I'm so proud of both of you. OK, great, well done. OK, no, we are working from a lot of uncertainty. So I'm just-- all evidence is interesting. OK, great. So now we are down to, basically, we're saying, OK, now there is 2,000 crores on the table. And this is, of course, Babik's now favorite topic. What happens to this 2,000 crores? How does it get split up now? And again, we are going to go through the same set of people 1 by 1. You're taking a 0, it's 20,000 crores. Oh, I'm so sorry, it's not-- I'm thinking about the subsidy, right? 20,000 crores. Yes. 20,000 crores, great. So let's go 1 by 1. Government-- oh, actually, let's start with banks. So it's the banks. And in the banks, there are two parties. The ones who receive the money, the acquirer, and the ones who are sending the money, right? The acquirer here gets the lion's share. It's about 60% of the transaction volume. He's the one who gets the money and then sends it down the road, gives another, say, 30%, 35% to the other bank. And the remaining 5% goes to NPCI. So as you can clearly see-- and this is, again, reading the tea leaves. Of course. I have no proof to the fact that this will happen. But where this comes to the apps is the banks, because the apps are getting the demand, processing the transactions, making sure customer experience is great, shares some money back with the apps. Now, in the credit card world, that used to be around, say, 10%. So I'm assuming this will end up in the same world. Sorry. Can you explain that? When you mean that in the credit card world, it was 10% who gets 10% in the credit card world? The network gets all the money in the credit card world in the banks and passes it back to the cards, right? Oh, of course. That is the app. And the app here would get probably anywhere between 10% to 20%, depending on how they negotiate. So this number that I was talking about, 6% then becomes 0.6% from the app side. And here is my favorite part of the debate. But the bank now for a very little loss makes a lot more money and becomes positive. And for the app, was a lot more negative. And with a little more money, still remains very negative. OK, pranav, do you concur? I think, see, a couple of things here. One, I think there's a distinction here between the consumer app and the merchant app. So my view, just like how there are two types of banks-- Yes, similar bank and other banks. --is your two types of apps also. OK, five, we're getting deeper. But once I was like, before we get to that, can we just talk about the headline first? I just want to make sure that between the split between-- OK, can we just assume-- Actually, let's go to the others. Consumers, we know that consumers still stays 0. Nothing changes over there, OK? Government, well, they stop giving the subsidy, I guess. Now, why should they give this? So if anything, they get more positive, right? Let's fare. I actually, to be honest, they won't stop giving the subsidy because somebody has to still incentivize the transaction under 2000 rupees. OK, fair. And so I think they've already signed off the FY27 budget. I don't think this is going to be easy for them to switch off, because I'm pretty sure the banks and the financial ecosystem is going to cry about the cost that they have incurred. So there are going to be two fees in the fare enough. So interesting. So now you're basically saying from a government standpoint, more or less they are where they are before. Nothing changes. And they continue to-- but anyway, we agreed that was a big green because of all the savings that they made from printing money and all of those other things. And plus, second order, third order effects, fare. So that's also a green. We'll come to merchants. Keep merchants on the side for a second. Or should we want to cover merchants? Let's talk about merchants right now. Merchants were essentially zero earlier, now merchants have to paste something. Yes. that moment. So the range that he was talking about, say, if it is around 30 to 40 bips, for the larger merchant, it materially hurts them. Right? I'd say, what does that number at about 25,000 rupees, it becomes like a hundred rupees, right? Of cost. And now at scale, this is something that they'd like to avoid. So they, and we can take it up now or later, but this might create opportunity to create new products, but this might also create an opportunity for them to change behavior. If you've seen in the past in different countries, where there are fees on credit cards, the merchant upfront tells you, hey, I'm going to charge you through a charge you extra, or you can pay me and cash. And so the merchant at the time of pass decides to change behavior just because there is fees involved in the ecosystem. Got it. So this additional 20,000 pros that comes into the system is almost completely funded. Completely funded. Yes. So let's say, so merchants are now 20,000 in the debt. Yes. In the in the red, right? Which kind of merchants will come to, but I'm just want to make sure that the merchants now have definitely been worse off. Now we are coming to this side. This 20,000 close suppose I asked both of you. I mean, of course, Bavik's position is very clear. But Bruno, suppose I asked you, if you're asked to split it between banks and these apps, which is the phone pays and Google pays with the world, is it fair to say that maybe in your telling Bavik, you're saying 90% goes towards the banks and 10% goes about 85% or 90% of the banks and 10% with the app. Bruno, do you concur this broadly? I think it would be a little bit higher for the apps. So my view is that about 30% would actually stay with the apps and about 70% would go to the banks. May I ask why you think that that's different from Bavik? Of course, Bavik we know is completely biased, but apart from that. See, I think a couple of reasons. So one, if I look at a debit card transaction today in the country, where a merchant pays say somewhere between 40 to 90 pips, the acquirer, which is the guy providing the card terminal, makes somewhere between 10 to 20 basis points, which roughly works to about 25%. So they get to keep 25% of the fees. So to me, the acquiring side that app, whether it is, you know, PTM or Google, Google P or any of these, would keep about 25% and there will be some fee that is earned by the consumer app as well, which is a app that the consumer is using to make the payment, which I assume would be about 10%, so net net 8 or 10% let's call it. So therefore, about one third in my view will be retained by the app. So it's not so bad. And of course, bulk of the money will go to the banks and that's that's the way it's going to be set up. Also, there's an interesting point here when Trump talks about the apps, right? All apps aren't equal because there is a skew in what kind of market share they hold. Very good. Because there is market share, there is negotiation ability, which is like while on the higher side phone, they might go and like push them at 30 or 40% a challenger app has no choice. But actually enough at 10% of 15%. We'll come to that. So now we know at a high level how this gets split up. So just to summarize, the new thing is consumers, nothing changes, government, more or less the same as before. They are what they were before. Now they are the same merchants down 20,000 crores. Banks, out of this 20,000 crores, let's assume it's closer to 80% of it goes to them and 20% of it goes to the payment apps. Okay. So now we know that banks are better off payment apps seemingly better off merchants worse off. That's really the new equation. Now we'll go into the details of all of those. Let's keep the banks on the side because banks find whatever. Is there a difference in the banks? I think it's going to get distributed relatively equal to their how much of accounts and consumers. Until now they were not incentivized to create better systems, better processes, you know, better overall infra, right? Now they are perhaps because they as a part of the fees, the merchants, especially the bigger ones are going to be extremely demanding. They want to say, hey, I'm getting charged on per transaction. I need x up time. I need y settlement time. I need z dispute resolution customer support like right now. So the overall experience of the platform increases. Now because there is an incentive. Exactly. And my assumption is because it increases at a platform level. The smaller merchants also get a lot of benefit from it, right? So this is, while they do lose money in the MDR, I think the overall infra effects just because the platform itself makes money is something that's definitely going to play out. Interesting. We'll come to that. Now I guess we are getting to the last two parts, which is the most interesting part, which is now on the payment app side and on the merchant side. Now let's talk about the payment apps. And this is where Bavik you have been basically saying, it's great for the big apps, like the phone pays and Google pays of the world and not so great for the challenger apps, which in this case is like the navies and the creds and the pop UPI and the super money, etc. Explain. So yeah, even so A, I think we have established that even on the large side, you know, a phone pay versus a pay TN, the ones who have over indexed on the merchant side of the ecosystem over indexed on credit seem to be making a lot more money. These are the pay TMs as an example. Yes. So this is the, this is the first point. The second point is, yes, I think and I was looking at some of the predictions from Bernstein and Jeffries, etc. Everybody is expecting the top three to make anywhere between 700 to 700 crores every year, right? And this might be the deficit that they require to plow back in probably gaining market share, right? Not just on the consumer side, but also on the merchant side, which is a very big deal. The reason it's a big deal is because on the consumer side, typically, economies of scale play out, right? I'm sending a payment to you through digital means, I can proliferate that information, I can use it as a source for growth, but this is not the same thing when it comes to merchants and acquisition of merchants. It is a very vertical, I vertical, it's almost like you go street by street. It's a very geofenced acquisition. So you can use a lot of it gain more. So basically, scale will concentrate more scale in my opinion. Interesting. Sorry, I want to talk about that number for a second. If you talk about 20,000 crores that comes into the system and even if you take the way you split it, which is that 80% of it goes to the banks and 20% of it goes to the payment providers, etc. So that's what. So 10%, so that's what. 4,000 crores is the pool that's available for the payment providers. And out of that, what 700 crores goes to the top three is one estimate. Each. Oh, wow. Okay. Those are the numbers that I'm just taking it with the pinch of salt. Those are the numbers that I saw and 500 to 700 crores each on the top three. Pranav, do you concur? It could be a bit higher for some payment. Okay. We'll be conservative here. Okay, fair. So we know that I get what you're saying. So you're saying that out of this money that comes to the payment apps, it's disproportionately going to the big apps. That's really what you're saying. Pranav, I think you had a very similar point of view when you wrote in the report. See, I think there are two or three different plays here, right? One is yes, the bigger apps will get more money because they have a greater market share today. But you need to see this as differently of the consumer side on the merchant side. On the consumer side, my view is while within the payment apps, the bigger ones might say, hey, we're going to take away the line share. It also creates an incentive for a lot of new players to jump in, right? So think about any digital platform which has a large user base. It can come up and say, hey, why should I pay a ATM or somebody else for consumer side? Let me just enable UPI payments on my own platform. Remember a lot of platforms got a payment license a few years back and then realized, oh, this is actually doesn't make any money. Let me give it back. Exactly. Now it makes money. So suddenly you might find your food delivery app saying, hey, I have a payment app as well. Sign up with me because I already have a customer base. I don't need to pay to get customers. So you will see an increased competition in the consumer side, which is probably what the regulators also want because you know, they've been talking about this whole 30% market share cap, etc, which they found it very difficult to pull off. Maybe providing some incentives gives more people a reason to go and try and win some market share. I actually agree with this point. So the challenges in this market will not be net new companies with who are starting zero, right? In fact, even if you look at the growth story in UPI, the top two apps have concentrated like over 80, 85% market share for the past five years. May 2026 is the first time it's fallen below 80%. And that's happened on the back of apps, which have one of two powers, power of distribution or power of capital, right? So you have like WhatsApp, UPI, you have super money with the Flipkart distribution of like 400 million. That happens because they have cheap access to users. And like I just pointed out in the unit economics, that is the bulk of the cost that you bear, right? Like getting users to your platform, figuring out how to engage them. If you already have like a mother ship who's solving that problem, then all you've got to do is like layer on other features and businesses on top. The other one like Navi falls into the categories that it had the capital capital as a mode on top of it, manufacturing as a mode. So they started off as a credit business, giving out loans they had acquired an NBC, they acquired a mutual fund house and they started giving off loans. At some point when they had reached a threshold where the cost of acquisition of a net new customer was high enough that their LTV didn't make sense, they decided to like go out and start to like build UPI as a capability. And now they are number four in the entire league table, right? So now they have figured out the entire play on how to target the right users for credit. Who are you pay users credit also is a behavior, right? In about 70 to 80% of the situations, it is at the moment of need that a customer looks for alone, right? But at that moment, they are going to look for things that they remember. So in that sense, like if you see a phone pay or GPS stickers all over your neighborhood, you are likely to like open those apps and search for credit. So how do people like credit or or Navi actually get people to take loans? It's because they get these users to form a behavior to open your app every day. And now that they open their app every day, they make them sort of like notice, oh, you know, I'm eligible for a 10 lakh rupees loan. So at the point that I needed, I'm going to want to take it from that ecosystem, which I now have built trust and you know repeated engagement over a bit of time. So that's that's how these ecosystems work. So now we went about going about it and watch happen, you know, super money on the other side. So these are the two big challenge or place. I agree with enough that what's going to happen now, again, because of the fact that it's an additional incentive. There is an additional incentive, which is the big merchants have to pay out money, which until today, I didn't have to like the likes of like Amazon or Flipkart. Even today, like process about 70, 75% of their transactions on UPI, right? Now that they have to pay out money on it, the obvious question is, why do I give it out to the ecosystem when I could just like be, you know, an affiliated party in this mix and become the UPI app myself? We spoken about Navi and super money and everything on this side. I think there is one player that we haven't spoken about, which is Beam. Do you think Beam falls into either of these categories or Beam is a category of its own, which is like a regulator has its source, but unless I would imagine it's also has a capital play. Is that what you think it is for Beam? Because Beam is also climbing up the charts. It is and I think it's the capital. It's a lot of money that NPCI and the regulator is making. It's plowing it back. But to be honest, I don't think the reason though, is to like create like a stellar play, like introduction of new features. If anybody in the UPI ecosystem were to like talk about their experiences with the regulator, they would say that the regulator actually comes out with features much faster than they can build them. Correct. I've heard this many times. So for that reason, I think Beam is like the showcase app. If you've got to use UPI circle, if you've got to use UPI. Here's a sample. And this is at scale also. This is not like a deba. And so you've got to like build it exactly like that. So I think Beam for lack of a better reason is the showcase of the ecosystem that NPCI has built around. So you said this is one aspect, which is about the we're still finishing the consumer app side. I think your broad point is that there are incentives now for challenger apps, even though the lion share goes to like the top apps, there is some now financial incentives now for others to start to create UPI apps of their own. If they have at least one of these big powers, if they are one of these big powers established, anything else on the app side on the consumer side, I think that's it. So consumer side, it was concentrated market share, no monetization. Now, I think it'll get slightly more fragmented with better monetization. That's how I see on the consumer side. Correct, you had a point to make on the merchant side on the consumer side. So you basically said that, oh, there is going to be different incentives now in terms of acquiring merchants for these apps, whether you are a phone payer, a Google payer, a pay team of the world. Could you like explain that? Okay. So let's switch to the merchant side. So the merchant side has been a tough business because it already generates money because people are charging. I mean, the platform, the charging merchants for sound boxes, smart speakers, positive devices and whatnot, yet it is not as, you know, as aggressive an environment as the consumer side, because you need a large feat on the ground, you know, a pay team, for example, has like 40,000 people on the ground to run this merchant business. And that's not everyone's cup of tea. And therefore, they have stayed away. There are two or three players battling it out. Now, what can happen is if the consumer side incentives are, let's say, on par with the merchant side incentives. And unlikely, but let's go to a hypothetical situation where that happens. And if there is a payment platform which plays on both sides, then making a lot of money on the consumer side will allow them to turn more aggressive on the merchant side. Whereas if you had a pure play merchant business, then there's a risk that you're not getting that windfall on the consumer side. And therefore, you're going to be at a slightly weaker position versus somebody who has both the arms. Now, this is again specific, if you're talking about how two or three players will interact in this new environment, then this becomes important. But if you take at an aggregate level, I think merchant business will remain a painful business for most people to bother with. And hence, I don't expect a lot of changes in terms of the cost of characters involved. It will probably remain a three or four player market, maybe the relative strengths change a bit here and there. But I don't expect a sudden flood of competition coming in just because there is an improved monetization. Let me just illustrate this as an example for our listeners. So, if you take Altex-Pacific names. So, if you take an example of, let's say, phone pay and pay DM. Now, phone pay is a app that is really strong on the consumer side and relatively weaker on the merchant side. I say, relatively. So, what that means is phone pay strength comes from having many consumers. And of course, it has merchants as well. But pay DM on the other hand is much more skewed on the merchant side, which means that it has many more merchants, which is acquired through feet on the ground as you both have described. And it's less on the consumer side. Now, what you're saying is, in the previous world, when the consumer side was like, you couldn't directly monetize. What would happen is, you knew you could directly monetize only on the merchant side. There was fewer incentives for a pay DM to go and try to acquire more on the consumer side. Because there are, look, we have the merchants. This is where the money is. Great. Let's stay where we are. Phone pay on the other side may want to go and acquire more merchants, but what are they going to fund it with? Because they anyway have a lot of consumers that don't directly give them money. But now, what you're saying is, because now the consumer side suddenly activates and gets revenue for a company like phone pay, they can use that capital to say, okay, let's now go aggressively on the merchant side and try to acquire some more people. And that makes the merchant side acquisition become much more, shall we say, competitive? Actually, if you, this might not completely be true. Because you think the money is so less that it won't matter? No, so one of three reasons. The first is that on the consumer side, apart from purely giving out cash, right? Cashbacks, the second largest growth strategy is an offline strategy. And this we have like validated several times a word. What does this mean? That means discovery of your post machine, your logos, your stickers in the offline world is a much better distribution mechanism than the next one. So it's putting out money as a play, putting distribution on the offline world. And the third is like using network effects like a referral and other plays, you know, to grow. So actually, the play that the ATM has been after, which is going heavy on merchants helps its consumer business grow pretty fast. It's something that's underappreciated, but it's largely true. Second is that on the on the merchant side, right? There are interesting and this is like a slight counter to what Parana was saying. If you see internationally, wherever there has been a regime change of introduction of a fee. There have been new businesses that have come in. Like if you consider us, right? The introduction of fees and a change in regime got the likes of Stripe Square toast and these businesses didn't exist before it. I think Brazil is also an example on something that has similarly happened to what's happening in India today. The large assumption here is that all merchants aren't made equal. So a jeweler and, you know, an NVFC and a school are not the same when it comes to their requirements on. Hey, I want to get settled on on T plus zero versus I'm okay being settling settling on T plus 30. Hey, I want like credit. I don't care about credit, right? So I have a feeling that on the merchant side, there is a big niche to be made on specialists, which on the consumer side does not exist. So there can be like on the consumer side, you're basically just asking people to come in, you know, scan their QR code or you're coming through the ecosystem, press a button, go pay and get rewards, right? And with so all products are sort of made equal, right? But on the merchant side, you could go really deep and think that, hey, I do T plus zero settlements best in the ecosystem, right? And build the right features for it or build the right business model for it, go and pitch that as your value proposition, right? Sorry, I would disagree. I think we, I think globally we're moving towards like large platforms on the merchant side. I think the specialization will just get eaten away by scale, right? Because you just work with the largest platforms. You can always be like a software service provider doing this niche thing, but when it, you know, let's have an education specific software, which also does payments, which would have just be white clabling somebody else's payment systems. But I think across the world, we've seen that you're seeing emergence of what I would call payment superstos, which does online offline, T plus zero, T plus 30, whatever you need, BNPL. everything gets, you know, house under one entity. I don't think you can beat this is the classic de-aggregation, aggregation playwright. Like for example, Bangalore has several startups right now which started off with the school payments as a mod. Got into like credit for schools, then got into credit for parents to pay for fees. And so they are creating this entire payment ecosystem around schools and education, right? Now the question is, is it ever going to be as large as a reserve pay or a pay you? Correct. The answer might be not. But that's exactly what tends to happen. So they have de-aggregated and gone after one vertical. I'm pretty sure that their entire VC pitch deck would say that now we'll do it for international education and then we'll do it for this. Because it's using VC pitch decks. Okay, hold on. Last point of the discussion, let's come to the merchants finally. The ones who are going to pay this 20,000 crores estimated, right? I guess that's the last piece of the puzzle. First question, what's the distribution of this? Is there like, is there a very similar skew like this? Like, okay, let me put this thing. Is it then fair to say something like say 70% of it comes from let's say the Amazon flip cards this weekies of the world? Yeah. The payment is the answer is yes. Yes. Wow. Okay, fine. All right. So let's just say so that is also very heavy in the peritocyte. So it means that there are a few people who give like a lot of this big share of the money. Yes. So if there are a few people who give a big share of the money, then it comes back to what you said, Bavik. Then it means that now foreign Amazon has Amazon pay and Flipkart of course has super money. But even for like a Swiggy or a Zomato and all of these other forms now, they are now setting up payment, they're dusting of their payment licenses and they're setting up UPI apps now, aren't they? But they have known this for a while and so they each of these companies for the longest period of time, like in the last four, five years have had their own Fintech departments and they have like independently hired, like consultants, et cetera, to just set this up like a financial services company have gone about building a lot of products in this space. So you have co-branded credit cards. No disagreements. Co-branded BNPL products. No disagreements. So they are selling loans. An example of that is Flipkart, right? Flipkart is selling at about a hundred million dollars worth of loans on a monthly basis now. Really? Yes. And so I mean through the ecosystem and the plumbing that super money created, but it's effectively on Flipkart, the app. So how can you now visibly differentiate it against a Fintech anymore? So a lot of them have known it for a while. I think the existence of an additional fee is just the sweetener. By the way, I must tell you that the first episode of Tuberto that we did was title can Flipkart become phone paper, before phone paper becomes because Flipkart was trying to basically use Super Maria and becoming exactly like a Fintech, while phone paper was trying to use pin code and trying to become more of an e-commerce company. So he's like, okay, who's going to come from there? But I think my question is your point is well made where you're basically seeing that look people have known this for a while. Keep Flipkart on the site because I feel like it's a bit of an exception. But if you look at say a Zomato and a Swiggy, et cetera, as an example, Zomato Swiggy and even say Amazon to some extent, right? These players now, you're right, they always knew it. But now they're going to feel the additional pain of a few hundred crores hitting them every month. Yeah, because the payment gateway or the payment speed is probably that hits your bottom line. Correct. And they hit that the most. Absolutely. And that's the reason that they would build it out. Absolutely. So now what are the second order third order effects? Now if you think that now if these people, let's say the top, I'm not even saying the rest. Let's take the top 10. The top 10 can easily like build out something. Now it's a different matter whether they'll be successful or not. But now the incentives for them are to say, you know what, this 20,000 crores, we are not going to give it away. If we take, if we give it, we will only be the ones to take it. That's really what they will drive. So I think it's easier said than done because if I use US as an example, other than Amazon, not many have gone ahead and built their own payment gateway. So I think the same logic exists in every market. Even Amazon actually have recently agreed to work with some of the external payment platforms. So I think it's not that easy. Again, goes back to that whole scale argument on the merchant side where I think scale benefits will trump any sort of specialization, etc. So it will become a bit of buying versus building and most likely buying would make sense for 99% of them. So I would doubt if it's more than two or three platforms which go ahead and build this out. That's that's one. The other is the other is that on the merchant side, I think online was already expensive, meaning that unlike the offline where it was really zero, online world, you know, you had transaction charges, platform fees, etc. that were being charged. So the platforms have had their time to assess whether they should build or buy and they have reached an equilibrium. So I don't think it gets upset by, you know, it's at 20bps getting added. I think the fees that were involved were reasonably significant earlier as well. I think the change, if anything, could be on the offline world where the fees is completely new. And here my view is none of the offline platforms would really go and build a payment app just to save off 20bps. And also they've always been used to paying fees. I mean, they pay for credit cards, they pay for debit cards, they pay for BNPL. The last point I'll make is if you remember a few years back, PayTM had come up with a BNPL product, right? So that's an even more attractive product than this MDR on UPI because PayTM used to make like 200, 300 bips on those products. But we didn't see a lot of people rushing and trying to build a BNPL product because they realize it's not that trivial to go and build it out. So I would be surprised to see new players jump in and try to become payment platforms on the merchant side. Consumer side, yes. Like I said, Zomato or Swiggy or any of these platforms which have consumers will say, let me also give you a UPI handle. Merchand side would be trickier. So what you're saying is that you can't expect now Chroma to suddenly spin up a UPI app of itself. Fair point. Yeah. I mean, a Chroma or a Vijay sales of the world. That is the world that we are talking about, right? But there are interestingly, there are some products that don't exist today. Like, for example, on the non commerce platform side, discovery of commerce offers is going to become like a thing. Right? Like, for example, assume that there is a travel service provider. Like, for example, like make my trip and make my trip run transactions with like several UPI providers. Now which UPI provider should I show on top is actually like it can also end up being like an ads business. Exactly. Exactly. Like some sort of a circular economy. I pay you money. You pay me that money back to show you on top. So I think that can become a big deal. On the on the commerce apps themselves to decide which any on the on the fintech apps who are processing transactions, you obviously will discover like commerce offers and that is how they will incentivize people to run transactions saying that, hey, you were already running this UPI transaction. Here is an additional incentive from the merchant and so this was like true for the credit side of the world. But this was not true for the debit side of the world. But now that there is a fee, I'm pretty sure that somebody is trying to like tell you, hey, I'm going to like undercut my fee. So show my offer on top for this transaction for this UPI transaction. I'll pay you like two rupees or five rupees, right? So that's like a so on both sides, there is like a lot of interesting products to be built. But there's a third site, which I'm actually fairly excited about. The existence of a fee now creates a requirement for somebody to manage the fee. Explain. So if I'm the ecosystem and I'm charging the fee, I'm definitely not going to help you manage the bad effects of the fee. Like for example, did I get charged for the right transaction? Did I get overcharged? Is there a dispute pending? So there's going to be this new ecosystem of like auditors, broad apps, AI agents to manage. I was waiting. We nearly got to the end of the episode without a mention of AI, but okay. So yeah, I mean, this fee will also attract tax for that matter, right? So there's going to be like GST, 18% GST on this fee. Who's going to like manage the refund with the tax department? So I'm pretty sure there are like some AI agent companies going to go after saying that. Hey, you don't worry. Whatever money that you're going to lose over here, we'll recover that money for you. And this is our payment. Actually, you're right. I just realized that if you're a merchant and let's say somebody makes a transaction on UPI, which is above 2000 rupees and goes through all fine, everything great. But then there is some return or there is some problem or there is some something that has to happen. Yeah, you're right. There is another before there was a fee there was no need. I mean, the reconciliation existed only to make sure that the money flow is cool, right? Now that I have to pay for that money flow, I'm going to be equally anal saying, did I pay more for this money than I was supposed to? And then there's going to be a new class of agent again, I start up trying to solve this. Fantastic. Really, really interesting. I think this is such a wonderful episode. Is there anything else on the new side that you want to cover, either with respect to any of the other stakeholders? On the merchant side, there are a couple of more thoughts that I had. As soon as, so until now, there was, especially on the offline world, the soundboxes, right? Soundboxes attracted rent, right? But rent perhaps is like only seen and I'm just speaking slightly psychologically as a way for you to have paid for the sound box correct so the merchants offline didn't demand for services as much right but now that you're actually paying pearl transaction You're going to be far more in a, so you're going to say that, hey, this, um, T plus one settlement that I was doing and you were charging me for T plus zero. Package that as part of the red or package that as part of your MDR charges or like get me like better, uh, dispute resolution or like help me like get some services that you were not offering me before. So I think got it. So there is a need for. So I feel that I'm paying my behavior towards like wanting value for that payment changes. And I think this is another space that you will see offline startups behave very interesting. Thank you so much, Prana Bhavik, such a wonderful episode. Uh, look forward to see how all of this pans out. Awesome. Thank you. Thank you, Praveen.