New payment tech unlikely to keep consumers from riding the rails
The desk believes that while the card networks face heightened regulatory and technological risks, their underlying business models remain robust, supporting continued revenue growth. Per the full note source, the card networks are projected to achieve low double-digit growth annually, driven by factors such as population growth and the ongoing shift from cash to electronic payments. Our consensus target for the sector is 1.075, with a range between 1.04 and 1.12, indicating a cautious but optimistic outlook amidst these challenges.
What the desk is arguing
BofA Global Research believes that recent de-rating in card network stocks is overdone, as structural growth drivers remain intact. The desk views legislative and AI-related risks as manageable, and sees BNPL as a growth opportunity rather than a threat. AI agents may change how consumers discover products but are unlikely to bypass existing payment rails.
Where it sits in our coverage
Internal coverage data is not available for this specific topic, as the article focuses on U.S. card networks rather than currencies. However, the consensus view across payments research is broadly constructive on card networks, with stable long-term growth trajectories.
How other firms see it
No specific firmIds are cited in the source; however, some research houses have expressed caution on legislative risks (e.g., Durbin Amendment expansion) and competitive pressure from fintechs. Others, like Goldman Sachs and JPMorgan, have maintained overweight ratings on Visa and Mastercard, citing pricing power and network effects.
Key takeaways
01Card network stocks have de-rated over concerns about legislation, AI, and BNPL, but BofA views these fears as overstated.
02BNPL is no longer an upstart but a segment with compelling growth ahead and an evolving business model.
03AI is more likely to enhance product discovery than to disintermediate existing payment networks.
Market implications
If BofA's view is correct, card network stocks (V, MA) may rebound as sentiment improves. The thesis supports the resilience of traditional payment rails and suggests limited disruption from new payment tech in the near term. This is positive for USD-revenue-dependent American Express and Visa.
Risks to this view
Key downside risks include: (1) new legislation capping interchange fees or mandating routing choices; (2) rapid adoption of BNPL leading to credit losses; (3) AI agents enabling peer-to-peer payments outside card networks; (4) regulatory action on data privacy associated with AI agents.
Hello, and welcome to Global Research Unlocked, where we discuss what's rising from growth industries to rising risks and opportunities in global markets. I'm TJ Thornton, Head of Product Marketing at BFA Global Research, and we're recording this episode on Wednesday, March 11, 2026. What maybe has gotten people a little over their skis, however, on the concerns here is that an agent on behalf of a merchant could put up a roadblock to the existing card-based payment or electronic payment and say to the consumer's agent or directly to the consumer himself, sorry, you can't pay with that modality.
We only want a lower zero-cost stablecoin transfer of value and no longer want a credit or debit card. Other views on the biggest payment networks and stocks have changed in the last nine months. Cyclical consumer fears have been part of this shift, but more than that, there are structural concerns around what AI and stablecoins could mean for the so-called rails.
Matt O'Neill, who recently reinstated coverage of the Payments and Processors Group, thinks that creates opportunity and he'll discuss that today. But even before these AI concerns were around, there were important shifts happening within payments, including continued growth in Buy Now, Pay Later, or BNPL. We'll get into that as well.
Thanks, Matt, for joining us today. Thanks, DJ. Looking forward to it.
Okay, the big payments names you cover, the brands we often see when swiping a card or holding our phones up for payment, these stocks are trading at some of the lowest premiums to the S&P on a PE basis that we've witnessed since the global financial crisis. Investors have historically liked these names because they don't take credit risk. There's a structural growth story for cards.
They're fee collectors. They've got high margins. But the multiples at which these trade suggests the thinking on these is changing.
Why is that the case? Yeah, absolutely. This is a great question and certainly one that has been dominating our conversations of late.
The card networks have historically been considered among the best and highest quality business models in public markets. They sit in the middle of every card transaction, as you alluded to, collecting a small toll on trillions of dollars of payment volume without ever lending a dollar or taking a penny of credit risk. They're inherently asset light, generate enormous free cash flow, have very impressive margins north of 50% and benefit from the continued secular shift from cash to electronic payments.
For years, the markets rewarded them with premium multiples, sometimes over 30 times. Today, the premiums on these stocks has shrunk to levels we haven't seen since the great financial crisis or at least in the last decade or so when compared to the market on an equal weighting. One of the reasons isn't that these businesses have gotten worse overnight.
It's that the risk perception around them has shifted meaningfully. There are really two big buckets of concern. The first is regulatory risk.
This isn't new, but it is intensifying of late. In the U.S. in particular, the focal point has been around the Credit Card Competition Act, which in a nutshell would require every credit card transaction be routable over at least two unaffiliated networks that break exclusivity with the major networks and introduce the threat of competition. If the merchants in theory could route around them, pricing power would erode.
When you peel back the onion, you start to understand that the economics generated from large co-brand relationships that are immediately earned by the card issuing banks are ultimately then shared back to the co-brand partners. Think large airlines, hotels and retailers. If the CCCA Act were to go through, it would very likely reduce those economics, thus reducing the value and the perceived value of those rewards programs to the consumer.
As we wait and watch what occurs here, the CCCA has been proposed to be attached to a couple of different bills. It hasn't made it through committees. As the legislators get the education that they need to understand maybe the second order consequences, the likelihood that the bill were to actually go through probably goes down increasingly over time.
At the same time, there is the other camp of risk that investors have been focused on with the networks for many years. That's the technological disintermediation risk. And that's more structural.
And this is where the discussion is focused around everything from AI, agentic commerce and stable coins. As you mentioned, people are quick to fast forward to a future where a consumer's agent is effectively negotiating with a merchant's agent. And in the background, they've come to a conclusion that is effectively mutually beneficial.
That future, in our view, is quite far off. One that would maybe say, hey, this merchant's willing to actually give that consumer a bespoke discount if he or she is willing to pay for this transaction, not over his or her existing credit or debit card. However, in the interim, I think where AI sees the most utilization is actually in the product discovery side of consumer commerce, whereby instead of maybe using a search engine or being shown various products in social media flows and consumers ultimately clicks on and ultimately lands on a merchant's website and buys the item through that route.
It's quite a long way out to think that the established trust network, the many-to-many problem that the networks solve for, and all of the minutiae details that are easy to forget but are critically important in consumer payments around fraud and chargebacks, returns, etc., haven't necessarily been solved for outside of the real model. There's very likely a long interim phase here where the status quo prevails. And it will take quite a while for there to be compelling value on both sides to affect the consumer behavior change in the value flow of payments.
Okay, first off, it's amazing that we're talking about agents conversing with one another, but that's for another time to discuss further. And as you said, that's fortunately a little ways off. Your point is that the business won't really get eroded.
But then on top of that, you do expect revenues of these companies to actually grow, and not only grow, but grow in the low double digits every year for the next few years, at least. What's driving that growth? There's probably still some cash-to-card conversion that's happening, but aren't we fairly late in that process?
Yeah, TJ, this is a great question. The investor community focuses in on this with great attention. The secular story, which is precisely that cash-to-card transition, I think investors and consumers inherently know in mature markets has by and large happened.
That said, if we take a step back, the growth algorithm for the card network model is still quite elegant and definitely worth walking through for a moment because there are some moving pieces driving it and also the underlying mix. If we go through them in order, there is the basis of personal consumption expenditure. As populations grow, more people spend more money.
That's your first baseline population and personal consumption expenditure that can be further catalyzed by inflation. The next is the secular displacement of cash-in-check transactions to electronic. That's largely, as you mentioned, a story that's played out in Western economies, large economies like the US.
There are examples around the world where either alternative solutions have arisen. We see QR code-based account-to-account models prevail on a domestic basis. However, those models still require a card-based issuance for consumers from those places for when they want to travel internationally.
But with the advent of the internet, e-commerce, and more and more spend happening online where you can't pay with cash or check, or at least in any easy, reasonable way, you end up defaulting to electronic forms. Rounding out the third and fourth points to the growth algorithm, their international opportunity is still very much alive and well as those economies modernize, which governments are usually quite amenable to. There's an opportunity to decrease gray and black market economies, as well as driving tax revenue as a greater percentage of local commerce is being done in an electronic and hence trackable way.
The final point, and I think this one is an important nuance, is that if we think of the networks at their core, they're helping to provide authorization, settlement, and clearing for card-based transactions. And that has been their core since they became public in the early 2000s. Over time, however, these businesses have grown to be much more than simply the rails for payments to ride on.
They've introduced through businesses that they've both acquired and built internally, very sophisticated fraud prevention, data analytics, and consulting-like services that are sold back into the market largely to the card-issuing banks, but also to, in many cases, large merchants. These are all bucketed under what's called value-added services. Those value-added services in aggregate have been growing north of 20%, which is roughly 2x the rate of growth for the core networks business and the high-single-to-low-double-digit range more recently.
We expect that growth to continue. The penetration of value-added services is not ubiquitous. There are large areas where the network businesses can continue to design, develop, and deliver these capabilities and affect positive outcomes with respect to lower fraud, higher authorization, and greater card utilization among consumers.
Got it. Just as a quick aside on that one, if somebody uses my card for a fraudulent transaction and I report it, I'm not on the hook for it. But who is?
Is it a combination of the bank and the merchant? Yeah, absolutely. That's one of the virtues that have made this such an attractive proposition both to the consumers as well as the merchants, that there is recourse here when there is fraud.
It goes back to the card-issuing banks and all of the players in the middle are actively inclined and usually compensated and incentivized to help predict and prevent as much fraud as possible. In the early days, if we think about fraud on cards, it was very easy to clone a card. The magnetic strip, any simple card reader could pull the data off that.
You could actually emulate that card and basically copy that detail onto another card. You could use those credentials in an e-commerce environment where the physical card wasn't even needed. That gave way to the EMV card technology, the chips that you see now, and has further given way to what I would argue is an even improved experience both online and offline by way of cell phone pay, where not only is the card being tokenized or effectively encrypted, that's what those EMV chips do, and that's what is so hard to emulate unlike a magnetic strip, but they further generally put a biometric in front of the consumer.
The old days of a nefarious actor finds a person at the back of a taxi and they take those cards on a shopping spree, that was fairly easy to do. Now, they find a phone, they can't unlock it, they can't use the payment wallet in that phone, it's useless. The fraud continues to improve, nefarious actors continue to get smarter.
A lot of what the network businesses have done is leveraged the data scale that they have and their ability to actually see and predict fraud. Oftentimes, fraud pops up in certain pockets and they're able to say there's a risk coming from this geography or this area to this type of a merchant and much more quickly tamp it down than was historically possible. The consumer's not on the hook, it's a huge value add to him or her.
The merchants are generally not on the hook, the bank issuers usually are the ones taking the risk on that as well as the actual underlying credit risk of the consumer, in fact, paying back in the case of a credit card transaction. This is an often underappreciated and taken for granted virtue of the established system. Makes sense.
That's a good segue to this next question, which is on AI agents and how they may potentially change the landscape. I could ask an AI agent to buy groceries for a March Madness party that I'm having. My school isn't in it, so maybe that party is to root on whoever is playing against Michigan.
Sorry for the Michigan fans out there. But would it be easy for that agent to pay with something other than a card? Maybe a bank transfer, maybe my stable coin wallet, and maybe the store is going to give the agent some kind of incentive to do that and so that they save on these fees.
As a consumer, I may not want to do that because it's extra steps and it's so easy to pay with a card, but maybe an agent could negotiate it, as you mentioned earlier, and could go through the extra steps in order to do that. Yeah, this is a great question. This kind of comes back to that longer-term future of agent-to-agent activities completely happening in the background, almost unbeknownst to the consumer and ultimately merchant.
This is where the legitimate concern and implicit overhang, particularly on the credit card networks, lies. There is some nuance here on both sides of the argument. Absolutely, there is a future scenario where that agent-to-agent negotiation could occur and there could be a solution that appears optimal to both agents that circumvents the traditional bank-issued card network rail of payment that we are all so used to and accustomed to.
If you think about the motivations for a human consumer, it's often speed, convenience, reliability, safety against fraud, etc., the ability to dispute a charge, get a return funded back to the bank account, etc. What maybe has gotten people a little over their skis, however, on the concerns here is that an agent, on behalf of a merchant, could put up a roadblock to the existing card-based payment or electronic payment and say to the consumer's agent or directly to the consumer himself, sorry, you can't pay with that modality. We only want a low or zero-cost stablecoin transfer of value and no longer want a credit or debit card.
The challenge here is today we have interchange-funded bank reward programs, particularly in the U.S. where interchange is still high and credit card rewards are still quite rich, that a consumer would be not pleased to be told he can't use that payment mode. In other economies or in a future where the CCCA did in fact go through and push down credit interchange, thus eliminating the value of those rewards, you could envision a scenario where a consumer may be a bit more agnostic as far as how they pay, not to say that they may be completely open and willing to change how they pay. Changing consumer behaviors, particularly in financial services, is quite challenging.
It's usually a bit of an uphill swim. Most people don't want to change their bank account or where their direct deposit goes or their card or preferred method of payment. This would be no different.
And maybe it's not as important for a grocery example, but for lots of other types of consumer payments where you want the ability to return a pair of shoes that ultimately were the wrong size that you ordered online or otherwise, there aren't great return channels, customer service channels when there's disputes or fraud or other challenges with shipping or otherwise outside of the current rails. I come back to the point that there may be certain unique examples where an agent-to-agent interaction can in fact drive a mutually beneficial outcome for both the merchant and the consumer. However, in a world, particularly in the U.S., where there are still high reward credit cards, that's going to be a couple of percentage points a merchant will have to effectively negotiate back to the consumer's agent or consumer directly by way of a discount or some sort of other reward loyalty points.
I think for that kind of long, medium-term phase, if you will, we're going to see the prevalence of AI much more on the product discovery, much more of a targeted product discovery assistant, and less so on an optimizing for maximum value to consumer and minimum cost of payment acceptance to merchant. How are the younger generations behaving differently than older generations when it comes to payments? Is Buy Now, Pay Later an example of something that's more heavily utilized by the younger demographic?
Do you think that will stay with them as they get older? In other words, they stick with this even as they earn more money and perhaps that's an issue for some of the credit cards or even the banks? Absolutely, TJ.
As we think about the generational dynamics around Buy Now, Pay Later, I think there's a couple of factors that all conspired together over the past decade that really gave rise to not just Buy Now, Pay Later, but broadly a collection of fintechs, if you will, across neobanks, Buy Now, Pay Later, and others that really were the outcome of a couple of things. If we go back to the great financial crisis, that younger generation that were still in middle and high school were watching the older generation, their parents, struggle with what was happening and saw credit card charge-offs and delinquencies and defaults and so forth occurring all around them. That gave that younger generation a healthy fear and skepticism of the established financial system, if you will.
Banks were introducing monthly minimum balance requirements and if you were too low on that, you would get monthly fees. Those monthly fees could be onerous and people were looking for alternatives that pushed them into things like no-fee neobanks. And for the access of credit, again, with the backdrop of an aversion to the traditional revolving credit card where you can build up a bigger balance, Buy Now, Pay Later really helped fill that void because it allowed a consumer to think about using credit on a single point transaction in a somewhat healthier way in that they knew explicitly what the payback period was.
A lot of the players on the short duration end of the spectrum started with a pay-in-four-like product where you basically pay a quarter today and then a quarter over the next two week periods or six weeks later, you've now paid off that particular transaction. That resonated with this younger demographic and allowed people to say to themselves, I can make this purchase today even though I don't have the $150 in my checking account to buy that pair of sneakers or whatever the item was. But I know I'll get my direct deposit consistently from my job two weeks from now and two weeks and so forth.
Six weeks in, I'll have paid it off. That younger generation who started using them, of course, has continued to get older. They start making more money, they're spending more, and even younger generations coming in behind them are also receptive to the Buy Now, Pay Later kind of consortium.
The longer term answer is that there is a healthy place between paying for something in full today using your debit card or cash or check and putting something on a credit card knowing full well that you don't have the ability to pay it off in an efficient way and it may become a persistent kind of revolving balance that you're only making minimum payments and ultimately being assessed a fairly high amount of interest against where Buy Now, Pay Later fills that sweet spot in between. Over time, however, that younger generation that we're talking about that started off using Buy Now, Pay Later in the infancy of that business model, they will get older, they will make more money, they will have financing needs that can't be done in a Buy Now, Pay Later fashion, whether that's an auto loan, a mortgage, things like that. I think that population will grow, but I think that they will also, by and large, graduate, if you will, as their credit needs become larger from a dollar amount and particularly around larger asset purchases where you need a mortgage, an auto loan, etc.
Buy Now, Pay Later is absolutely here to stay. Will it completely disintermediate the established card players and banks anytime in the near future? No.
At the same time, the banks are putting features in place. Many of the large banks within their credit card programs provide consumers an option to take single transactions on their credit card statement and actually turn them into an installment-like loan. Okay, got it.
Last question. You mentioned how traditional banks are getting into that Buy Now, Pay Later business in some cases where you could take one of your larger payments and instead of paying big interest, you could divide that payment into four and just pay it off. What about the Buy Now, Pay Later companies expanding into what the banks do?
You could argue that they've already done that to some extent, but I'm talking about things like loans where maybe they're extending more credit than they have historically. Are they doing that and do you view that as a positive or is it a negative because it makes these companies more credit-sensitive? Yeah, that's a great question.
I think the answer is somewhere in the middle. When the Buy Now, Pay Later wave started, the focus was overwhelmingly on the shortest duration, so the pay-in-for, much lower-ticket type transactions. Part of the virtue of that for the providers was that while they were implicitly providing credits to a consumer on a transaction, they were really not a consumer finance business in the classical sense.
I think the challenge is that many of them have evolved into more of a hybrid model or in the process of evolving into more of a hybrid model where they are pushing that duration and that average dollar per loan higher. As you do both of those things, you introduce non-linear risk into the model. The probability that somebody pays back a $100 transaction over the next six weeks is quite high.
The probability that somebody pays back a $3,000, $4,000, $5,000 transaction over the next three years is significantly lower. Again, going back to the genesis of the short-duration low-dollar Buy Now, Pay Later providers, these companies weren't really underwriting the consumer. They were really marketing a payment alternative.
When somebody didn't maybe pay back that first Buy Now, Pay Later transaction, they just simply wouldn't let them make another one. They were much more marketing businesses than they were credit and consumer underwriting businesses. The proof will be in the results going forward, particularly for those evolving upstream, if you will, from that sort of low-dollar, short-duration to higher-dollar, longer-duration model.
Finally, as it ties back to valuation, the concern is always that the longer duration your loans are, the more concentrated they are, the inherently more cyclical your business becomes. Much of what has kept multiples on traditional consumer finance, credit card, and broader financial stocks lower over time is largely attributed to that cyclicality. Things are good when things are good.
But when things are bad and unemployment takes fire, delinquencies increase. They require much larger provision expenses, which dramatically weigh on earnings. It's that inherent cyclicality that keeps investors from ascribing too high of a multiple to those stocks.
I think there's going to be different examples and a balancing and potentially for some a reckoning of those who've gotten too far over their skis thinking that they could easily swim upstream, go into higher-dollar, longer-duration, and having realized in hindsight that maybe they did so a little too aggressively or irresponsibly. This will be the next couple of years as these companies continue to grow, not just in transaction and volume, but as the underlying makeshift goes towards that higher-dollar, longer-duration. There will probably be a bit of a shakeout between those who are in fact quality underwriters doing this responsibly and those who maybe have been a little bit too risky in the interim.
Okay, Matt, thanks very much for the time today. It was an excellent discussion. Thanks, DJ.
I really appreciate it. The traditional credit card rail system works well. And though merchants pay fees, some of those fees are returned to the consumer in the form of rewards.
Some of those fees go toward covering and preventing fraud. AI agents in a science fiction way could threaten that. But even there, both consumers and merchants may be loathe to accept stablecoin or direct payments.
In order for those transactions to make sense to the consumer, they'd have to be compensated to offset what they'd otherwise get in rewards. Then the advantages for the merchant may not be all that attractive anyway. The CCCA or Credit Card Competition Act could threaten this whole ecosystem.
But that also comes with a bunch of unintended disruptions and consequences. And that's a reason Matt thinks support for this could ultimately win. Thanks for joining.