Harshita Rawat: Visa, The Next Decade of Growth

[MUSIC] In the know with Bernstein Research. Welcome to In the Know with Bernstein Research. In this series, we discuss investment controversies together with what is top of mind and in the news with Bernstein's research analyst, who are in the know. Our disclosures can be found at the end of this and every episode. I am Diana Wood from Bernstein's Boston office, and this episode features our senior US payments analyst, Harshita Rewan. In today's episode, Harshita digs into the investment case around Visa. Visa has underperformed the market by about 10% year to date. Multiple's are down versus June since the stablecoin narrative took hold amidst the passage of the Genius Act in July of 2025. And we did lower our global card volume forecast for the first time in a decade earlier last summer. Many developed markets now already have a high-car penetration and in places such as India, the opportunity is increasingly harder to realize. And yet, we feel just as good if not better about Visa as a business. It is easy to attribute the underperformance to the stablecoin narrative, but we wonder if the bigger deal is a combination of funding sources for the AI trade, get reclassifications, and rotation within financials. But on a fundamental level, our confidence around Visa is better now than it was five years ago. It is controversy around the name, however, and who better to walk us through it than Harshita. So without further ado, welcome, Harshita. Thanks, Diana. We're so happy to have you here. Thanks for having me. Okay, so let's jump right in. When we think about the traditional Visa business model, we think about, you know, global credit card penetration. So what is the credit card penetration now globally versus the US? And how do you feel about estimates for that number? Yes, Diana, when you think of Visa, you would think of this rising tide of digitization, right? Like cash to card conversion, fueling its business. And some of those trends are starting to mature as you alluded to. Global card penetration today is 65% back in 2019. It was 48%. Now while there's still plenty of cash left to digitize, even in developed markets, such as continental Europe, places such as the US and UK, or starting to enter a mature phase. US card penetration, for example, is now at 78%. It was kind of 67% back in 2019. So the card volume growth globally has decelerated, right? Like to 8%. It used to be 10% historically. It'll slow even more to 6% to 7% over the coming years. So for an investor, it used to be quite easy to have a bull thesis on Visa five years ago. Just look at that global card penetration number that I cited earlier. And historically, you could easily get that 10% revenue growth based on that GDP growth and cash digitization layered on top of it. The story is now harder. Given some of the maturity dynamics I alluded to earlier, two things, and I think where investors may be missing the forest for the trees, amidst all of this, one, there is still robust growth runway in the number of transactions, which now drives more of the revenue growth for Visa. So basically, what is left to digitize is this smaller ticket transactions, which are numerous. And there's also this opportunity in new flows. And the second point, which I think people don't appreciate enough, is this growing portion of Visa's revenue growth is now coming from value to services, which are now 29% of Visa's revenues growing high 20s at this point twice as fast, if not faster than the core. Wow. OK, so let's double click into that a little bit. So value added services or a VAS as the acronym says are, as you said, increasingly becoming a growth driver for Visa. But before we get into that, what are value added services for them and what is included in their portfolio? Visa's perhaps the single most frequently asked question I get from investors now. One of the reasons why we see this opportunity in Visa's stock kind of derading to 20 times 2027 onings for a double digit revenue growth is this kind of, I think, confusion around what value added services are for Visa. So they're basically an array of processing security fraud solutions, data analytics, quieter services that Visa is now increasingly selling to its customers. There are dozens, if not hundreds of different kinds of value added services. And because of this lack of unifying team, it has been harder for investors to forecast this. Now, some examples for these services would include token services where you have an encrypted version of your card credentials stored in an watching environment and they could be a bunch of services wrapped around that, issuing services, think of card loyalty, benefits, acceptance solutions, think of much in processing gateway risk solutions, fraud solutions. And then also some consulting and advisory services here. Okay. And does it also include, it's so funny I was traveling with my children the other week and got like fragile in charges on my card when those pop up. So they literally that is what's part of this VAS is when they let you know, absolutely. I mean, there is such high degree of fraud, right? Like that could happen in digital transactions, as you can imagine. And by the way, and I know if we're going to talk about a genetic later in our conversation, it's going to proliferate even more with a genetic and visa. Remember, Visa and MasterCard combined, they look at more than 10% of global GDP within their transaction data. That's the amount of data they look at. So they have a lot more insight into fraud trends and they can really wrap it around in a value of services both for merchants as well as banks. As well I know. And over, I have to imagine, sadly, over the last decade or however many years that fraud piece and you think about the context of cybersecurity has to have grown like specifically on its own. Absolutely. It's a big headache for banks and merchants. And in fact, if you think about the only years of e-commerce fraud was rampant, right? And I think then Visa and MasterCard came up with a bunch of these services, both on tokenization, both on kind of advanced authorization, etc. Which really brought down a lot of the fraud in e-commerce transactions. And for merchants, it's not just the fraud piece, right? It's also the charge back piece. But what we are also seeing is friendly fraud with a lot of transactions where the cardholder actually did conduct a transaction but they disputed it for a variety of different reasons. So I think merchants also increasingly look for these tools for to also combat friendly fraud. We had friendly fraud. I hadn't really heard about this more. So people just actually buy something that they go and dispute it and get the money back? Exactly. And this is a big headache for merchants and I think Visa and MasterCard, because they look at a lot of the data, I think both these companies have made acquisitions on this charge back management for merchants, which is becoming a bigger and bigger deal for large online merchants. Wow. Okay. So what makes the value added service growth runway durable? And how do you think about the TAM, sort of the total addressable market for this piece of the business? Yeah. I think we can somewhat size the addressable market for value added services at $500 billion of revenue. The current revenue base for Visa and value added services is a little bit north of $10 billion. And then this is spread across advisory services, risk security solutions, some of which we talked about, some issuing solutions, some kind of acceptance solutions on the merchant processing side. But the thing is that there are just dozens, if not hundreds of different kinds of value to services. So it kind of becomes a little bit more challenging to get more granular on the addressable market. Now look, VAS runs across a customer journey from acquiring a customer, onboarding them, the actual transaction, managing post transactions. We talked about the charge back and fraud. And then also engaging customers. I'm sure you see a lot of card linked offers, for example, these and master cards can also do that value added services. I think a good framework in which we think about the growth of value added services is that two-thord of VAS is transaction linked. So you do get the natural equation of this transaction growth, which is high single digit. Remember, it's a bit higher than the dollar volume growth numbers we talked about because we're increasingly digitizing smaller tickets transactions. And then you layer on penetration of value to services per transaction. As Visa puts value to services into its go-to-market engines, Visa, for example, now has enhanced go-to-market for value to services. Now, but 450 specializes people didn't have that before. And then there is a big chunk of value to services, which is not transaction-related. Right? I think about marketing services, consulting, data, solutions, open banking, which both support core engagement, enhances the mode around the core. So, for example, things like consulting and marketing services. But then also as this diversified revenue stream, Visa, and this is something people don't appreciate enough, they oftentimes would have existing commercial agreements, kind of, I think a seat at the table with the largest financial institutions around the world, the largest fintechs around the world. So sometimes it's also a matter of bundling different products. And also, Value the services also helps them to exercise some pricing because you're adding more value on the transaction, right? One final thing I'll add here is over the past year, we have seen this new visa as a service construct, kind of this unbundling of the visa network, to support networks outside of visa. So think of that as like domestic bank-to-bank payments beyond visa score. And this also presents this value to services croat opportunity. This visa as a service construct beyond visa as a money movement network. Okay. So what about, and we can speak more to this, but what about disruption risks? So I focus a lot of my time in equity sales on the international side of research within Europe and obviously within Asia. And we hear a lot about pay-by-bank in emerging markets. What do you think about that and others as sort of disruption risk for visa? So this is a key topic for investors, Diana, as you can imagine. Disruption of cards on surface, you know, cards feel quite dry for disruption. Credit card fees that merchants pay in the US, for example, are often 2 to 3 percent, although debit card fees are much lower, you need to make that distinction. There is always talked about pay-by-bank, stablecoins, fintechs, encroaching upon the networks. All that said, though, pay-by-bank has gained traction in some countries as you alluded to. India is an example. I do think that conditions in these markets were quite different versus the US to begin with, to lead to this kind of growing adoption of pay-by-bank. In markets such as the US, we have seen limited traction of pay-by-bank. Some of this could also be an opportunity, this kind of pay-by-bank growth in some countries for visa to layer its services on other networks, at visa as a service construct we were talking about. We have already seen this kind of being put into action, for example, with what they are doing, with layering some fraud solutions on account to account payments in the UK and Brazil. Remember, Apple Pay perceived as a threat for visa and master card 12 years ago, that has been a gift that has kept on giving in terms of digitization of small value transactions for visa and master cards, which has massively helped with the sketch to card conversion journey that we have been on for the past decade. I should have sent this originally, but let's just go back in time a little bit. But for listeners here, can you just explain what pay-by-bank is? Yes, absolutely. So pay-by-bank is basically if you're using your bank account to make a payment directly to a more chunk, or some other party, it is functionally similar, if you think about it, to a debit card payment, right? Debit card essentially is doing the same thing, the only difference is when you make a payment with a debit card, it goes on visa's rails. If you use, let's say, a pay-by-bank, let's say, UPI in India or PIX in Brazil, you're using the local rails for that transaction on a debit card. So increasingly more, and in emerging markets, this is, well, I mean, it's not necessarily a disruption threat, but people are using pay-by-bank more. In some specific markets, like in India, we have seen a lot of adoption of UPI. In fact, in India, debit card volumes are getting decimated, because of UPI. But an interesting fun fact, credit card volumes in India are growing double digits, and vast majority of Indian card volumes were visa and master card or credit cards. So people still like to use cards in these markets. Debit card value proposition is what is under attack, for example, in India. But then, if you look at markets such as Europe, debit card interchange, which is a big bulk of the fees that merchants pay for debit cards is quite regulated. It's very low at 20 basis points. So we've actually seen a good solid growth of debit cards in Europe. So it's a very market-by-market phenomena. One thing I always like to say is payments is a very local market. You cannot kind of draw broad global kind of conclusions. You need to look at market-by-market geography by geography. Yeah. It's very fragment, in a nine-sense, I feel like. So what about stablecoins? Which I mean, crypto or stablecoins, just digital currency in general, I feel like has been the topic digital for years now, increasingly so recently. But what about stablecoins? What do you think about them first? And then how do you think about the customer value proposition around that? So I'll say something counter-induced of your stablecoins offer card networks like Visa, new opportunities to make money and not lose volumes underlying. And strategically enhanced their networks. So for example, through faster settlement, Fiat to stablecoin can have conversion capabilities. I do think there is a good opportunity for new flows. So think of business to business, business to consumer, remittance says, which is something which is not typically card it, so not something that Visa and Mastercard kind of go after. So there's some opportunity with stablecoins and those flows. And right now, as we sit here today, Visa has more than 100 stablecoin linked card programs. The stablecoin linked card volumes are more than $4 billion annualized for Visa at this point. Visa is making money in stablecoins right now. Now, I guess the big picture question that I often get is Visa kind of right for disruption from stablecoins in the future, right? Like a terminal value concern that the people always can ask about. Now, I guess if I take a step back for a second, payments, disruption of stablecoins will not be in areas you would think, right? First of all, vast majority of stablecoin activity today is in crypto capital markets, not in payments. So you really need to when you look at any stablecoin numbers out there, you need to kind of look at it more granularly. Now, in retail consumer to business payments, the ones that cards kind of address stablecoins, in a way, are a solution looking for a problem for the foreseeable future in our view. In most developed markets, there is this huge chicken and egg problem of building a merchant acceptance network across 100 million plus merchant acceptance points and a consumer network across billions of consumer cards, right? Debit, which is functionally similar to kind of stablecoins, right? Like moving money within your own accounts credit is a very different value proposition that's the money you don't have is already regulated and cheaper versus credit. So when people talk about this, stablecoins being cheaper than 2 to 3% guard fees, that's not even an apples to apples comparison because to the 2 to 3% is a credit card fees, not a debit card fees. And then cheaper rails, for example, in the US already accessed, right? You think about bank-to-bank that we were talking about and how it has had limited traction in the US. And I think when you talk to merchants and I do have a lot of conversations with merchants, you do wonder whether in a world where conversion for merchants online is a huge problem, right? Like a lot of e-commerce cards are getting abandoned. To what extent merchants would want to steer customers at the point of check out towards stablecoins where the conversion right now is very, very low. The consumer value proposition for developed market customers don't really exist. And then one final point I also want to highlight is card payments wrap many services around it, right? Like we talked about chargeback protections, fraud, mitigation, other kinds of value services. stablecoins just offer you money movement. Money movement is a commodity. The services, the governance you wrap around it, which is what Visa does. That is the actual differentiating factor here. So what we have seen ironically is the best consumer use case of stablecoin that we have seen so far is a stablecoin linked Visa card, which is growing extremely nicely. And that gives a lot of these stablecoin companies also monetization models. So for example, you are a Latam Fintech. You are doing a lot of remittances in stablecoins. There is a desirability of a US dollar stablecoin in those markets. How would those customers spend those dollars that they're going to receive? You need some kind of acceptance mechanism. Visa card gives you that acceptance mechanism. So we have seen a lot of those stablecoin linked cards. All that said though, I do think stablecoins have a lot of other use cases kind of which stronger product market fit beyond retail consumer payments. So you think about less liquid remittances. Starting to see stablecoin activity, cross-border business to business, treasury cash management. And there's also a greater need for on and off-ramp from stablecoin to fiat world that Visa and Mastercard can help with. Mastercard meet and acquisition here quite recently. So again, I think this is not a disruption risk to volume. This is an opportunity for Visa and Mastercard to make money in this orchestration layer from fiat to stablecoin world as well as near flows. Okay, this was so helpful. But at the beginning of your answer, which I appreciate, you were saying I'm going to say something that's controversial. Stablecoin, I'm then assuming, has represented part of the overhang on visa. Clearly, they're investors that are not understanding this hidden, let's call it hidden, but value proposition that it presents. Is that correct? So, Dan, I think we are also in a market where sometimes narratives are being developed to fit a stock performance. The stocks have underperformed in the past year or your today. Some of that is just those border macro factors that you alluded to. There was this whole kind of gix reclassification. This happened kind of two, three years ago, but it has the fact that visa and mastercard are now in financials. So, there's certain dynamics around that. There's a funding trade for the AI trade. And then there's a little bit of overhang from Stablecoin. I don't think it's a primary overhang candidly. And I think what eases some of these overhang is, I think, as you start to see more and more partnerships, more growth of Stablecoin linkage, more partnerships, more investments at visa and mastercard are making here, then you start to realize that this is an opportunity. A very interesting dynamic that I am seeing in my conversations with investors is the payment specialist. So, the folks that who are looking at this day in and day out and who have kind of done very deep work, they're not as what about Stablecoin disruption. In fact, they believe this is an opportunity just like we do. I think it's the generalist where there is, I think, some convincing to be done. I feel like that's so true. And my heart goes out to generalists, PMs, or diversified in various areas because it's within so many sectors, you have to pull back another layer of the onion to really understand the story because the overall narrative could seem scary or that this is the next disrupting trend and they kind of link on to that. And we are living in an environment where most major businesses that they're looking at are getting disrupted by technology. So, I think it's a very natural question to ask. But I think when you look at it a level deeper, what you realize is that visa is not about money movement, right? They can, they always use quantities bank-to-bank networks to do settlement. It's really about the governance that's validive services on the transaction. And can anyone else develop that? Yeah, sure. You know, you can develop those validive services. It's not that technologically complex, but how you can enforce it? That's the biggest challenge, right? How you can enforce it across tens of thousands of financial institutions and make sure all the rules and standards you're laying out are being followed across industry. That is the hot bar here. And you were just saying, you know, technology is rearing its head and disrupting so many aspects of our world right now. And obviously the leading star in that movie we are watching is AI. So, let's dig into that ever-popular topic. You know, what do you think about agentec payments and how are you thinking about AI in the context of these businesses? Yeah, Diana, this is a topic that we have been spending a growing amount of our time on. And the conclusion that we have reached so far, again, this is an evolving team, right? Is that currently agentec commerce is very much in an iterative phase with its product market fit and not ready to take over the full and often very asynchronous shopping journeys of consumers yet. What we have discovered through surveys, through talking to merchants and platforms is consumers are increasingly using AI tools for discovery as much as 40 to 50 percent of the consumers are actually using these tools for product discovery, but they're not yet using it for transacting. They have maintained relationships with their favorite marketplaces, their favorite retailers under 1 percent of web traffic can be attributed to referrals from AI chatbots today. So, it's not as high as you would think, although it could be a little bit higher for the DTC brands where it could be double digits. But I think more broadly what we're seeing is this grand vision, right? That was talked about six months ago around agents buying things on our behalf autonomously is looking a little bit far out, especially for medium to high consideration categories. Instead, we're having a lot of discussions with merchants, with platforms, with consumers around AI enhanced conversational commerce experiences, particularly around improved product discovery, not product transactions, but product discovery. And on both the leading AI apps, so you think about obviously charge, GPT Gemini, but also merchants owned properties, like their websites or apps. In fact, OpenAI is evolving playbook here, including the recent abandonment of Instant Checkout, which was just launched six months ago and they now abandoned it. Really highlights the difficulty of changing consumer behavior. Consumers do like to shop in merchants properties because you can get like a better vision of their product, more reviews, etc. And also the importance of integrations and underlying infrastructure and the challenges with checkout itself. For example, around conversion, OpenAI's Instant Checkout saw very poor conversions for some merchants, right? Like a lot of checkout was getting abandoned. There's a whole array of issues around post purchase. How do you kind of sing real time inventory pricing, merchant catalogs, etc. And I think as the conversation has still to towards discovery from AI, the futuristic scenario of delegated agent to agent transactions does look a little bit further out versus what we thought a year ago. The idea of the shopping piece kind of just as inside makes me so sad in this world of AI. Like not only do I want to be on a website picking out exactly what I want. I'm still the one that's shopping in the store talking to every girl in there like with my daughter having her try out. Like that's just part of the human experience. I could maybe have an agent like buy my toothpaste and paper towels and like I don't even know serial whatever that I just get. I mean my kids whatever toilet paper, but the idea this to me is just oh what are we doing? But it's great. But what are we doing? No and I think we have seen, right? Like for some I would say low consideration purchases. There's already an experience for that which is Amazon subscriptions which work really well. Right. Totally. Literally it just comes like it's on schedule completely. It's just yeah it's fascinating to me. But I think you know you're hearing an important point which is consumers like shopping. I know people like the shopping journeys and also shopping journeys or asynchronous. Right. So for example you discover a product most consumers don't buy that product then. Yeah. They come back to it later. They think about it. You know something else kind of I would say influences their imagination. They're kind of you know inspires them and then they come back to it. So it's very asynchronous. So I think this agent is kind of you telling an agent to buy you a dress and the agent doing it like who would want that? I know. I can't. It's also so this is like a whole different topic. But like what are we going to do with this unlocked time? I'm just like as humans. I mean of course and you're like a wonderful working mom and I am. I mean I we always need more hours in the day but like if you get a shop. Right. Yes. I'm going to shop and do a million I think. But if you take all of the like little nuances of what make life life away and you streamline and make your life that efficient. It's like what are you living to do? I mean I guess we could like hike a lot of mountains and like meditate and you know that's wonderful. But I just it's just funny to me like what are we going to spend our time doing? Anyway. Okay so my last question because we could do this for hours and I can't wait for my next pod with you all ready because there's so much to discuss. But I would just say in the broad scope of commerce like and I think you already answered this but will these payments sort of help or hurt the mode overall? Look I think we were talking about earlier how e-commerce in the early days was was rifled fraud. Right. Then if that was the issue with e-commerce then you can only imagine what happens with the genetic. There's going to be a lot of fraud. There's going to be a lot of disputes. All that means is more tokens for these and MasterCard more value of services around security, consent, programmability, dispute management. They could also be potentially more transactions if you're splitting one basket into multiple types of items right from a different merchants. And there's going to be so much complexity around intent. So just to give you a simple example if you ask an agent to let's say buy your nine-year-old son a car what if the agent buys a natural car instead of like a tonic car right. So how do you capture that user intent? How do you capture disputes like if there's a dispute who is liable for that? Is it the agent, the merchant, the consumer? How do you kind of establish an audit trailer for transaction identity authentication? So what this means is visa and mastercard will have so much more opportunity to layer their value at its services around this agentic transactions. And in fact, what we have seen over the past, I would say six to eight months is, yes, there's been a lot of protocols on agentic, right, like from Google and OpenAI and Stripe and also visa and mastercards. Some of them compete with each other, some of them kind of complement each other. But what we have seen is the entire ecosystem with Google OpenAI's Stripe merchants have kind of recognized that they need standards around liability and risk. And the networks are the best positioned here because they are the common standard center. And it was quite notable recently that Stripe, which has developed its own kind of agentic commerce protocol with OpenAI recently expanded some of their kind of capabilities to enable access to network-led agentic tokens. So kind of like a big win for visa in that regard. And the networks themselves have adopted their kind of networks and the tokens for the agentic world. And they may get richer data on those agentic transactions, so which could be quite valuable for their value to services business. So I think this could be the Apple Pay moment for visa and mastercard all over again, right, like how 12 years ago everyone was debating whether there's a good thing or a bad thing for visa, Apple Pay. And this was one of the best things that happened to them. And I think agentic is something similar. I kind of love it. They're creating a whole new market for themselves. First of all, the car analogy. I mean, that literally would be me. Tom, we had a birthday party for Tommy and COVID. And it's like, what do you do that's safe with just our neighborhood. Tom rented a like convertible card that out, you know, we could drive kids around its outside and say, whatever, I would be the one who would go online to buy him like more cars now and get a Maserati by accident. That literally would be me. That's going to be only just I I literally. And it's funny even right now. It's so it's just nothing as perfect. I was in Florida for Easter and I'm the bunny. So so much stuff was being sent. And then everything else I ordered for the week ahead was sent to Florida by accident because it's like you have to change the address. Like there are so many things that are imperfect about the processes in tech and then I can see continuing that you write. This is just unlocking even more value for the name. This is wonderful. Yes. Absolutely. Okay. Well, we do have to cut it here just because we could do this forever, but I cannot wait for our next podcast together. This has been wonderful. Thank you so much. This is so much fun. Thanks, Dada. You've been listening to In the Know with Bernstein Research. If you like the podcast, don't forget to like and subscribe. In the know with Bernstein Research. If you do not have access to Bernstein's research, you can find it at Bernsteinresearch.com where you can also find important disclosures that we encourage you to review. Bernstein has no obligation to provide any updates or changes at any time in the future. All references and or market forecasts are correct at the date of recording. The views and opinions expressed in this podcast are those of the presenter and may not be the same as the views of Bernstein or its affiliates. Bernstein is not providing any financial, legal or tax advice or recommendations in this podcast and this should not be considered as investment advice. This podcast must not be copied, distributed, published or reproduced in whole or in part. None of us hold positions in any of the equities that we have discussed today.