Circle Internet Group, Inc. ($CRCL)

All right. We are continuing here at Canaccord's Sixth Annual Digital Assets Symposium. And up next, we are very, very pleased to have our keynote session of day 2. And that keynote is with Jeremy Allaire, who is the Chairman and CEO of Circle. Many of you clearly know who Circle is and know the investment case here. But for those very few that may not, Circle is one of the true juggernauts in the broader digital asset ecosystem, and it is the leading global issuer of compliant stablecoins. The company's USDC stablecoins are true digital dollars, tokenized and fully reserved and which can move across blockchains and the international instantaneously and pretty much frictionlessly. This is a major distinction and upgrade from electronic money, which has been with us for a very long time and hasn't changed much in 40 years. We think of stablecoins as a true programmable money layer for the Internet and where honestly, the TAM is broader -- where the TAM is really honestly a broader M2 money supply over time, which in the U.S. stands at $22 trillion. Against this backdrop, we like Circle's competitive positioning in what might be one of the last major addressable markets that has not been truly disrupted by advancing technology over the next -- over the past 10 to 15 years. There are a lot of big companies out there looking at launching stablecoins. Some may come from the e-commerce world, some may come from social media, some may come from TradFi. But all of these players have existing businesses and in contrast, Circle is a pureplay in stablecoins with no ulterior motives other than USDC and the broader stablecoin market. So therefore, we view Circle as the Switzerland of stablecoins, which I think makes them a #1 choice when people are contemplating adoption of stablecoins. In contrast, and I'm going to ask Jeremy this in a little while here, would Amazon really want to accept a Walmart stablecoin? They may have to if they're compliant, but it doesn't seem like a natural partnership there. So I could go on for a while. But with that not so short introduction, welcome, and thank you for being with us here today, Jeremy.

Thanks, Joe. I'm very, very happy to have the conversation and be with you guys for this event.

Great. So maybe we'll just do a 2-minute quick intro, and then we'll jump into some more detailed questions. Intro on Circle?

Yes, sounds good. Yes. So I'm Jeremy Allaire, Co-Founder, Chairman and CEO of Circle. And Circle got started 13 years ago at really the inception of blockchain technology. And we have -- we were animated by a vision that over time that the technology of blockchains would allow us to build a protocol for dollars on the Internet and this would ultimately be protocols like we have for the web or for software delivery or for communications on the Internet and that would become possible that we could take dollars, fully reserved dollars and not fractionally reserved dollars and allow them to work on the Internet in an open, interoperable way and ultimately drive the cost of storing and moving value down to effectively 0, kind of commoditizing payment utility. We spent a number of years building towards that, working with governments, policymakers, regulators around the world, working with banks, other leading financial institutions to build up that infrastructure and that led to the development of our flagship product, USDC, which launched 8 years ago. And USDC had its start in the digital asset markets and is now really expanding into a really wide array of use cases from cross-border settlements for enterprises to infrastructure inside some of the leading payment companies of the world to being used as collateral and capital markets transactions in trading markets in various types of e-commerce transactions. And more recently, we're seeing USDC and related infrastructure being used by AI agents, which have started to proliferate even in the recent quarter. Circle has always been focused on being trusted, transparent, compliant, having good governance. That has set us apart. That has allowed us to work with the world's leading institutions. And we're -- we've been the first company registered and regulated in many, many markets around the world. The last thing I'll say with this intro is over the past 1.5 years, Circle has really been broadening out our platform. And so today, Circle really is an Internet financial platform company. We're building operating system technology with Arc, which is an operating system layer for the future economic and financial system on the Internet. We have a digital asset layer, which includes not just our stablecoin network with USDC, but it also includes EURC, the world's largest euro stablecoin, USYC, which is one of the very largest tokenized treasury and money market products in the world and other infrastructure. And then up the stack, we're building applications that benefit from all this. So Circle Payments Network, which we call CPN is an application utility that's being used by over 50 financial institutions and growing to create a compliant, easy way to move money globally using the power of stablecoins as a settlement technology. So we've really broadened out what we do, and that's a little bit about the company.

I think it's more than a little bit, Jeremy. It's a great technology platform. Most people just see USDC, but the ecosystem is big. It's providing a lot of people the ability to build and to make their own use cases from USDC. And it feels like you're just starting on that journey.

That's for sure.

Yes. So I just wanted to go back and I think for the benefit of people, and this kind of took like a minute for me to sink in when I was first learning the Circle story, and I have a fintech background. The difference between digital money, programmable money and what we call electronic money today. My analogy is if you use your PayPal or something on a website to do e-commerce, you're still drilling a portal back into old kind of established antiquated payment systems, even though you're on the Internet. It's really just kind of virtual representation of old school money. It's not really programmable. It is just what -- it was a nice little step forward. But I think that is truly not what stablecoins are. That's truly a programmable money layer. I'd just love for you to kind of provide your own differences between the 2 for people.

Yes. I mean, look, I think most of what we experience as sort of this electronic money or software-based money today is sort of lipstick on a pig as we say, it's sort of a user experience, a decorative user experience on top of an existing old plumbing. And I think when we started Circle, we really started taking first principles. How do we have a dollar, the safest form of dollar that we can come up with, which would be something that's as close as possible to like a government obligation dollar. And how do we actually represent that as a digital asset, that is a digital bearer instrument, like digital cash that can free circulate on the Internet. So really getting to the base layer of money and expression of the base layer of money where that expression has a native Internet infrastructure layer behind it. And this is similar when we sort of had -- people had -- if you remember, satellite radio was digital satellite radio or your cable video-on-demand was digital cable or there are 500 channels of digital satellite. It was digital, meaning it happened to use binary. But when we really think about what digital media was, it was when the actual physical manifestation of a song, of a movie, of a document actually became things, artifacts on the Internet that can run over open protocols on the Internet that have the total reach of the Internet and ran over open networks. That's what made those different. And that's really what we saw when we started Circle. We could get at that base layer and then from the ground up, build up other aspects of the financial system. And that's where the programmability came in. I think the insight that we had 13 years ago -- when we were getting started, there was this idea of programmable money. There was this idea that blockchain networks would become computer networks. They'd become computing platforms. They'd be operating system like technologies that didn't exist then. The idea of a smart contract was sketched out on a napkin at that point in time. But I had spent my career working on a lot of different Internet technology stacks, including programming languages, virtual machines, application servers, core infrastructure. And I could see clearly the power of a compute engine that could run on these cryptographic computing systems. And what that would allow for over time is a composable programmable financial system. You'd have a base layer of ultrasafe digital cash money and you'd have the ability to write software that could build higher-level abstractions, whether that be a loan, a trade, a bond, all the kind of financial primitives that are there. And we've seen versions of that really happen through DeFi. We've seen software machines that create options markets, derivatives markets that enable people to issue and tokenize credit and other things on these models. But I think really, we've reached a point where this was, in many ways, like an early adopter technology, experimental, but obviously still sizable. And we finally got, I think, the kind of legal and regulatory clarity around the world, not just in the U.S. that's now making this a kind of more hardened part of the global financial system.

That's great. That's a great overview and explanation. I like the kind of digitization analogies to other assets that have been truly digitized. And I think that programmable ability is something that's exciting, especially, as you mentioned, at that intersection of potentially agentic AI and stablecoins, which we are going to get to in a minute. But I also wanted to just at a high level, kind of ping you on what you think the TAM is here. I mean, if you think old electronic money is M2, and this is a newer version of M2 because it's really programmable and it's just as reserved, I mean -- it is money that's fully reserved. Do you see M2 as the TAM here? Is that the TAM? Or is it a piece of M2? Or how do you view that?

There's a few different pieces that we look at. And again, you have to kind of expand across the full suite of what Circle is doing to capture that. So the first is global money supply is, I think, around $120 trillion in value. And you have M1, you have M2. And what's interesting about it is most of M2 is credit. It's credit money. It's essentially kind of leveraged credit money. And I'll come back to that because I view that as part of the TAM. There's a huge portion, and it's estimated around $60 trillion of that money supply is in the form of either physical cash or noninterest-bearing demand deposits. So essentially payment money, working capital money in accounts all around the world. Stablecoin money immediately most significantly addresses that $60 trillion, right? So working capital money, payment money that exists in the world, but a far more high utility version of that, that runs natively on the Internet. But when we think about that total TAM, whether it's the $120 trillion or the $60 trillion that's there, there's sort of the money stock itself, which is one dimension of the TAM. And then there's all of the different utilities that utilize that. So those utilities include all of the exchanges that we have in the world. They include every payment system that we have, every payment utility and payment application, that includes the card networks or other payment schemes, B2B payment infrastructure products, all the different utilities from capital markets infrastructure all the way down kind of transform, move and use that money and extract fees. And so within the world of the utility that gets wrapped around that money, there's also an enormous TAM. And so our view is that more and more of the money stock will become truly digital currency native, and that will just grow over time. And more and more of the utilities for how that money is used will become software native, will become Internet native and will be delivered far more efficiently globally, et cetera. And so it's both of those together that really represents the TAM that not just Circle can fulfill, but the entire ecosystem that works with this technology can fulfill. Like there are going to be lots of different players that transform, whether it's the NASDAQ moving to tokenized equities and blockchain-based systems and sort of changing the way in which equity capital markets work to utilize stablecoin money, tokenized equities, et cetera. So there's going to be a lot of transitions that happen just like we've seen in whether it be enterprise software moving to the Internet or communication services moving to the Internet, these kind of migrations of these established utilities to more Internet native delivery. So the TAM really encompasses both of those and that -- and I think that's significant. And then for Circle specifically, we're playing in the application utility space with CPN. We're building a generalized payment network utility that is designed for financial institutions to utilize that. We're building the money layer across a number of different core monetary assets, key reserve currencies, treasury money, kind of T-bill money, and we're building operating system technology. And I think this is one of the most exciting areas, which is -- there is this new paradigm of economic operating systems. And these are an evolution of blockchains to provide an environment where economic activity can be expressed, executed, coordinated in a tamper-resistant way. And these economic operating systems will have significant fee revenue attached to them and other monetization opportunities as well. And we're going very aggressively into that space with the development of Arc. And so we're going to play at this fundamental infrastructure layer. So that's part of our TAM in the actual digital money layer, which is the money stock of this and at the application utility layer, which we think can be significant over time as well.

That's great. Updates my view on the TAM. So that's great. Thank you, Jeremy. So yes, so we've got this money layer, and we've mentioned Arc and CPN a little bit. But what -- stablecoins can move around on other blockchains today. Clearly, Ethereum holds a lot of the stablecoin volume globally. But what -- but Arc is an L1 that's being purpose-built, I believe, to drive transaction volume. It is an infrastructure layer of what could be a much larger base of economic activity. Can you just go into the vision on Arc and how it potentially evolves over time?

Sure. So the first thing I'd say, just to connect to one of the things that you were noting is that our stablecoin network today operates as of today, I believe, on 32 different blockchain networks. So we are the most portable cross-platform stablecoin network in the world. We've prioritized interoperability as a huge thing. And that means if there's a new network with new users, with new applications, with new utility, we want to make sure that our digital dollars and our protocols work there. And so we continue to do that, and that's been significant, and we work with all of these different ecosystems. And so that's one piece. At the same time, as we've engaged with the market over the past number of years and as major -- we've engaged with major institutions, global systemically important banks, clearing houses, global custodians, the largest asset issuers and asset managers in the world, the firms that run payment systems like Visa or Mastercard as well as governments and government regulators. What we have found is that the kind of early adopter phase of these blockchain networks was not necessarily suited to the needs of the mainstream scaling phase. And that there's a whole set of things that those types of institutions really need if they're going to migrate their activity entirely on to these networks. And so we looked at that, and we saw that Circle as a kind of market-neutral infrastructure company that's learned a lot about the use cases are for real-world financial activity. We were in a very interesting place to be able to go into this market. So both through organic R&D, multiple acquisitions, we've built this up. We launched Arc in test phase in October with over 100 major companies, including Goldman Sachs, Deutsche Bank, Visa, Mastercard and so many other incredible companies. And we're moving towards what's called Mainnet launch, which is the kind of full commercial launch. And I think for us, though, the vision is as an economic operating system, these networks are going to need to support a lot of different activity. So what is that? Well, of course, it needs to support the movement and transactions of things like digital dollars. But these also need to be kind of economical platforms for firms that want to issue digital token-based versions of their own assets. It needs to be a canonical platform for a bank issuing a tokenized deposit for an asset manager that's issuing tokenized funds, for equity issuers, for credit funds, for so many different forms of assets that exist in the world. So we've built fundamental infrastructure at the core that is for those types of issuers. It needs to be run as a network that is not a decentralized anonymous system, which is the way most -- excuse me, the way that most of these networks have been run, it actually needs to be a network where the firms that operate the network are known that they're held to very high standards of information security, compliance, operations, et cetera. And so the network is a network where the operators of the infrastructure are known, and those operators are ultimately serious financial institutions, serious financial infrastructure companies. It's a very different model than the kind of decentralized, let's build something that is away from government. Here, we're working with a global base of firms who can operate with this and that can uphold the standards that a major banking regulator would say, yes, you can put real money and real transactions on this. So we've taken that approach, which is different. And then the longer-term vision here is as more and more forms of assets and financial contracts move on to these systems, this will extend more deeply into corporations themselves. Just like corporations went through the process of being an off-line company to being an online company where they rewired their business processes using software that faced customers on the Internet, there was a whole transition that took 10 or 15 years, we're going to go through the same transformation. Corporations are going to move from being sort of online to being on chain. More and more of the things that are economic in nature, the treasury management of the company, the contracts of the company, the economic contracts of the company, the -- all of the financial activity of the company, the governance and voting that's involved in companies, all of those things are going to move on chain, and you'll have more digitally native corporations. And increasingly, and this touches on the Agentic and AI side of this. Increasingly, these corporations that move on chain, the primary units of labor are going to be AI agents. We expect that the vast majority of work that is executed in the real economy is going to be work, in particular, in certain industries, but it will be work conducted by AI agents. And those AI agents are going to need economic tools, they're going to need to have the ability to make and receive payments, you're going to need to be able to compensate external AI agents that are doing work on your behalf and have contractual arrangements with them. So this whole migration to an economic Internet is what we're going after with Arc.

That is -- that's a great explanation, and I learned a lot right there, Jeremy, myself. And I think that goes back to your earlier comments that you're a technology platform. Everyone is kind of focused on USDC and what USDC in circulation is right now. But that is -- I've been looking for, okay, how does TradFi really jump into the world of digital assets? Like is it clarity and then they're going to move in? How much can the big juggernauts in the sector move the sector forward by themselves, yourselves, other big players like Coinbase, doing a great job, but everybody else is not there yet from the TradFi world. And so it feels like that explanation you just -- or the illustrations of what Arc could be well beyond just applications related to stable -- your stablecoin, right, to a much broader tokenization effort of a lot of things, right, is like the first thing I've seen in a while that is, okay, this is a real tip of the spear for a broader economic migration. So that was really great to hear. So on that note, I mean, it's an ambitious. I mean, clearly, you've got -- you have the blockchain expertise, you've got the regulatory expertise, you've got the stablecoin expertise to envision and implement something like Arc. Is there -- are there other Arcs that people are working on that you're aware of out in the world that are similar in terms of their breadth and depth of capability to move the world towards more of a tokenized and agentic world? It's pretty interesting.

Yes. I mean, look, I think a couple of thoughts. I think the first is that the evolution of blockchain technology, in some ways, mirrors the evolution of things like mobile. If we -- I use that as a reference point because we can all remember, for example, like 15 years in the desert where -- like there was a new mobile platform coming out like every year, there was the PalmPilot, there was the compact iPack, there was the Windows phone, there was the Symbian phones, the BlackBerries, the NTTs, Smartphones, like all this stuff. And we watched as that evolved. And everyone was always so excited about the promise of mobile. And -- like 100,000 people get together every year in Barcelona, go to Mobile World Congress and they'd be all showing off like, look, you can do a mobile game or look, you can do -- you can like -- you can use this for your inventory tracking or whatever it was. And the reality was it was total c***. None of it worked. It was a really bad experience. The tooling, the user experience was poor. And these were huge companies, Nokia, Samsung, NTT, Microsoft, pouring billions of dollars into this bet that these platforms were going to become significant, right? But it really took -- there were multiple things that happened simultaneously, including iOS, including Android and really a convergence of technology forces that had to do with kind of capabilities of hardware, had to do with capabilities of software, how that got put together, and you hit this inflection point. And you all of a sudden had kind of an open stack that really enabled a surface area that people could innovate on and people didn't even know what you're going to be able to create. That was the amazing thing is that we went from like all these ideas about what mobile could be to this explosion of millions of apps that were there. Now I think we're sort of going through a similar transformation right now. And so to answer your question more directly, of course, there are others that are trying to think about this next generation of economic operating system that takes blockchain technology from an early adopter phenomenon to a mainstream scale phenomenon that penetrates the entire global economic and financial system. The opportunity is enormous. And I think when we think about Arc, we really think about it -- in some ways, it's a little bit like our Android, right? We're bringing it into the world. It's an open source technology. We're building the fundamental references, but then we're bringing in a huge ecosystem around it. We're bringing in the ecosystem across all the different categories of firms that would benefit from it. And we ultimately want them to be stakeholders. We want them to participate in governance. We want this to be a real ecosystem play that's open and value creating for everyone who participates. And so I haven't seen a project with the same sort of scope, ambition strategy approach, but that's not to say there won't be. I think in this area of agentic, I think there are a number of people, whether it's Circle, Stripe, Coinbase, Solana, others who have very quickly seen like, oh, this is here, it's happening, it's fast. We've been working on stuff in this space for a while. But now it's kind of -- it's caught a groove as we've entered 2026.

That's great. And I like that surface area analogy. Arc is that big surface area that was original iOS and Android, and no one knew what we were going to do with it yet, and it was definitely a lot better than that Palmpilot module for one specific application. People just should have been better off using paper, but everyone tried, right? So I do remember those days. So maybe we just kind of drill down a little bit on agentic. We've mentioned it a couple of times, but I think this is a nice area to talk about it because this Arc surface area is large. And I think there's probably been some early days experimentation with it. It does feel like a lot of the big players in the space see this also as an opportunity, right, to drive things like payment volume, bring cost down, expand distribution. So just kind of how it works and -- but it clearly makes a lot of sense that if you have programmable money going back to stablecoins and USDC and you have an operating environment where agentic activity can occur and then you add programmable money together, it's a really nice combination and foster a lot of innovation?

Yes, that's exactly right. I think there are a lot of pieces to understand this. I think the first relates to what you just said, which is the building blocks. So what do these networks and forms of money provide? They provide a globally available, interoperable open system that software can interact with. They provide a layer that introduces kind of a proving system. This is really important. We need trust in AI, right? We need to be able to have assurances that the work that's being conducted is valid, that the transactions that are being performed are valid, that the identities that these AIs are representing can be proven. And so we need these systems of kind of provers, which is what cryptographic networks give us, right? So there's like an underlying material there that's really, really important to AI, which for many people, one of the biggest concerns sort of is trust. Now it's also at a very basic kind of economic and technological layer that this is valuable as well. Hopefully, I don't have a cold.

Bless you.

Thank you. At a technology level, in a world of AI agents, there's one vantage point, which I think is kind of a looking-backwards vantage point. And there's another vantage point, which is a looking-forward vantage point. The looking backward vantage point is, I'm a person, I shop online, I want an AI to shop for me, and I want to empower it to go make those transactions on my behalf. That's looking backward, okay? That's useful. That will be something that happens. But that is not the big opportunity here. For what it's worth, I think that credit cards will be stored and credentialed and AIs will have permission to use your cards and things like that. I think with the proliferation of stablecoin money and digital wallets that support stablecoin money, that will increasingly be utilized in those transactions over time as well. But again, that's -- in my view, that's a looking-backward view. The looking-forward view, which is the one I'm most excited about, is that AI agents, when you look at the phenomenon that's happening now with AI agents and you look at the work that's happening with Claude Code, Codex and OpenClaw, what's happening is that the units of work, the units of labor in corporations are being broken down into agentic tasks and workflows. And the services that one might need to deliver something are being expressed more and more in AI agents. And you might have a legal AI agent, you might have an accounting AI agent, you might have a content creator AI agent, you may have a wide array of specialized tooling that is AI accessible and that is itself AI-based, and that whole system is going to involve value exchange. That whole system is going to require that AI is interacting with other AI in conducting transactions. And when you start to get into that world, there is no other infrastructure in the world that can scale from a micro transaction to a macro billion transaction, right? There's no system in the world that can do that other than blockchain native stablecoin models. And that's very, very powerful. So the bet that we're making is that the number of agents that are active economically is going to grow into the billions and the payment volumes that are involved when you have agents conducting very high-frequency transactions. Just think about this. If I'm an agent and I am going to go to another agent to conduct some work and that other agent is going to conduct work. And effectively, it's the amount of tokens that, that other agent consumes plus the value-added markup, whatever the margin is on that because of the value-added work that's happening there. That compensation might be work that costs $0.25. And the frequency of those transactions might be in the hundreds of millions of billions over time. And so you need an economic medium to do that. And so we're building that. We're building that at the core. We're building and contributing to the standards that make that work. So it works interoperably across a lot of different systems. And I think that is where we're going. And so I think people don't typically look at the world through an exponential lens. It's very hard for the human brain to do that. We look at the world through a very linear lens at best. Many people can't even go linear. They're just like, now the world is what it is and it's just going to keep going like what it is. Linear feels like, wow, it's a lot of change. Exponential, we just -- our brains can't even go there. This is exponential technology right now. And the acceleration that's happening from agentic is exponential. So when I think about a year from now, 2 years from now, 3 years from now, I imagine huge amounts of software machines, AI agents operating at scale in the real economy, economic operating systems as the backplane for that, stablecoin money as the value storage value transfer medium that works around the world. And when we think about -- if you're someone who's covered payments and people talk about what's your TPV numbers, what's TPV look like, what's your take rate look like, all this sort of stuff, I think a lot of that goes out the window. And TPV in this world will, in my view, likely be many, many orders of magnitude larger because the cost of transacting is approaching 0. Last week, we announced something called Circle Nanopayments. And Circle Nanopayments is in a beta test right now and it creates a way for agents to be able to store value in a unified balance and to be able to -- using a protocol called x402, which is a protocol that allows agents to conduct transactions with other services on the Internet. It allows that stored value, that USDC stored value to be spent to any kind of app across about, I think, a dozen different blockchain networks and can be spent in less than a second with the transaction costs that can go as low as $1 million of a penny. And so if you're talking about micro transactions with micro transaction costs that are basically from pools of value that AI agents are utilizing to interact with all these different services, that's the kind of infrastructure that you need. That's an infrastructure that will go live when Arc goes live. And so it's sort of a core primitive in the Arc infrastructure that we believe will be very helpful to all these agent deployments.

That's a lot to unpack, Jeremy. And I've been an analyst a long time, and I honestly haven't heard something as potentially compelling as that vision in quite a long time. So super exciting. Yes, my linear brain is kind of thinking, okay, this move forward with some genius was good and then clarity is good. My linear -- oh, yes, we're going to grow a little bit here and there because of these things. But it never really occurred to me that it could be -- it could just be AI that drives blockchain adoption and tokenization instead of the other way around, the tail wagging the dog here, dog maybe AI and the tail could be blockchain instead of the other way around?

Yes. We -- I'm a big believer as a kind of technology futurist and a technologist in general that -- the most exciting things happen when you have multiple compounding exponentials. Like that's when it gets really interesting. If you think about the iPhone, there was sort of -- yes, like hardware innovation was there, but it was sort of the adoption of cloud infrastructure on the back end, which was sort of the virtualization of what it took to deploy software. And it was the adoption of mobile broadband, so the CapEx spend that went into broadband, wireless broadband...

You have 4G for the first time to make it work, right?

Right. When you put together the hardware and the software, if you didn't have the cloud infrastructure that virtualized how fast and easy it was to build the back end of things, and if you didn't have the physical infrastructure at the wireless level, the radio level, if you didn't have all those together, those are like 3 compounding technologies that created the exponentials and explains also why prior versions of mobile didn't work. The hardware wasn't there. The capacity of the networks on the broadband side wasn't there. And it was hard to build services on the Internet. Cloud hadn't really emerged. And so those compounding kind of exponentials are where you see things kind of have these different kind of liftoffs. And I think AI, crypto networks and blockchains, regulatory frameworks for how this can be used, these are all kind of connecting all at the same time and are creating, I believe, these kinds of compounding exponentials.

Right. The setup looks good. I agree. No doubt. there's so much to talk about, but I think maybe we checked the AI box a little bit here and so the vision here is compelling. So Arc, you're going to kind of open it up to big established players, your partners. Does it go to full live? Or is it incrementally live? Because I mean, the vision here is so -- it's ambitious, and it feels like it's more than a flip of the switch for the whole thing, right, because there's a lot going on, right?

Yes. I mean, there's a whole sequence of things, right? So as in Testnet, we've got very strong developer traction. We're working with all of the different players in the digital asset ecosystem to make sure that their products and services, whether you're a custodian or an exchange or a wallet or a developer tool company that your products are compatible with Arc. And so we're optimistic about having that kind of day 1, like turn it on and everybody can use it kind of thing. The other is how do we make sure as an operating system, what you care about are what are the apps that are there on day 1. And so we want to make sure that some of the most popular apps that exist on chain today are there on day 1. And so we're working hard to make sure that we have that. And then there's the -- getting the kind of financial infrastructure partners to stand up and operate this and getting that -- going through a kind of live testing and operating cycle with those companies operating the infrastructure and essentially allowing anyone in the world the ability to test out a node to battle test the nodes, and that's an important step here as well. And I think additionally, it's getting new things that are starting to come into the market like these tokenized real-world assets and use cases that are relevant to larger traditional financial institutions, getting those to be able to be live when we launch or shortly after when we launch and things like that. So that's the body of work that we're doing there, and we're making really good progress. And obviously, as more and more rolls out, we'll communicate that out to the public.

Very exciting, no doubt. So I'm going to go back to my linear brain and some more simple questions here, Jeremy, in the time that we have left. In my opening remarks, I kind of talked about the stablecoin wars. And if Amazon doesn't want to accept a Walmart stablecoin, I mean, if they're compliant, I guess you have to, but how do you -- is that -- maybe that analogy is too simple. And then relative to your position as kind of the Switzerland of stablecoins, do you see enterprises TradFi, e-commerce guys succeeding here? Or does their established kind of big business kind of just create resistance from their competitive cohorts for broad-based acceptance?

Yes. I think for us, and this has been something that I think we've sort of been consistent about, which is that stablecoins are network and platform businesses. They exist as network utilities. Those network utilities have platform characteristics, meaning developers build on the protocols of those networks. And when developers build and integrate their applications to those networks, it creates utility for the whole network. And so it's a classic platform developer flywheel model. And that creates very significant barriers to entry. And so when we think about those for Circle, we have a very strong competitive position. We have the competitive position of having done the work of becoming the regulated infrastructure in all of the markets that matter around the world, and that's critical. We've done the work to build the financial infrastructure by working with leading financial firms around the world to plug in liquidity between global markets, regional banking systems and our network. We've built the flywheels from a developer perspective, and that started 8 years ago and has grown where now if you're a developer and you're building a product and you want to support stablecoins, if you don't support USDC, you're at a competitive disadvantage because it's so interoperable, so many products supported. And so that's what you care about. You want your users to be able to -- if they have digital dollars to be able to transact interoperably with as many people as possible. So we've created these very, very powerful network utility flywheels and those compound for us. The vast majority of implementations USDC, we've never heard of. They've never talked to us. There's no business partnership. It's just they implement it because they need to because it has such reach. So you have that effect. Additionally, you have what we call liquidity network effects. Liquidity is so fundamental to this. Users need to know that they can create and redeem at scale anywhere in the world, and they can get it in and out of currency systems, financial systems all around the world and that the actual pipes, if you will, can carry a lot of value. If you don't have all those things, you're not building something useful. And the marginal value of something that comes in that doesn't have those things is effectively 0. The final piece is that Circle, as you introduced in the beginning, we're a Switzerland. We're a market-neutral company. We're not an e-commerce company. We're not an exchange. We're not a brokerage. We're not a consumer payment application. We're not a payment service provider. We're not a bank. We are a market-neutral infrastructure and all of those types of firms are our customers. They all build upon us and can achieve their own goals by building upon us. As soon as one of those types of companies tries to introduce their own stablecoin, they're immediately in conflict with all their competitive peers. And so the incentive to adopt that other product is very low, notwithstanding all the other things that I just talked about in terms of what those barriers look like. And so it shows up in the numbers. I think this was actually part of what we shared in our Q4 earnings was it shows up in the numbers. Over the last year, there have been many dollar stablecoins that have been introduced, regulated products. But when you actually look at the liquidity, when you actually look at the transactions that happen, transaction volumes continue to grow. We've seen 250% year-over-year growth. There's some recent data out from third parties in the last week, which shows continued growth in transaction volumes. And USDC has continued to take market share. We're now the majority of transactions happening on chain, and that's been very strong growth to get there. But when you add together USDC and Tether in terms of actual transactions are happening with stablecoins on these networks, it's more than 99%. And so if you add together all the other stablecoins that have launched and all of their transaction volume, it rounds to 0. It rounds to 0. And so the point here is that these network effects are strong. The utility is strong, and we feel very good about our position. And I think for us, there are going to be people that experiment because they think they have a big brand or they think that this is something closed loop that they can do. But at the end of the day, general purpose, market-neutral kind of public Internet infrastructure utilities for digital dollars are going to be a winner take most market structure. and we're well positioned in that. And I feel good about from an investment perspective, what we're doing with the continued expansion of the liquidity, the availability, the continued types of major institutions that are integrating to our network, the advancements that we're making down the stack with Arc, up the stack with CPN, all of these things continue to contribute to the strength of our position. And so we pay attention to everything that happens. And obviously, I think when we were going through our IPO, one of the things that we talked about with you and others is in a winner-take-most-market structure, there's typically like 3 to 5 players. And we think we'll be one of those. We hope we'll be one of those. But there are going to be players that don't even exist today, right? They're not even in the market today. And so 3 to 5 years from now, there will be others, but we hope to be the most of the most.

Yes. I mean -- and I remember you saying the winner-take-most strategy, that was 9 months ago now. And no one else has really popped up yet. So the winner-take-most analogy is clearly still material. Maybe it's winner-take-most plus a little bit more potentially. We're almost out of time. Can the big banks make the pivot on this technology? Technology is hard. The banking philosophy and mindset is not necessarily incredibly innovative. They usually default in their technology to compliance and things like that. Do they cry uncle and just white label or partner? Or can they really kind of be part of this winner take a good chunk of the market at this early stage, especially when you think of this agentic layer getting added and what that may mean to acceleration and adoption of potentially more USDC?

I mean, my view is across the full spectrum, not just of using stablecoins as a payment system, but across the opportunities to offer financial assets and products on chain, the opportunities for banks are enormous. In fact, today, at Circle's global headquarters, we were hosting an event called Circle Current, and we've got roughly 100 top executives from the leading banks of the world and leading asset managers of the world. They're coming to engage with us around all of the different use cases that can be built out. And I think whether it's the SEC, the CFTC, the prudential regulators, every single one of them in the United States is giving green lights. They're giving green lights that banks can go, they can implement this technology. There's more and more guidance coming out. There's a set of guidance from the SEC, which is in draft comment period amongst regulators, which will likely come out. The GENIUS Act as coming online. All of these things are happening. And so we are seeing banks and capital markets players all gearing up to build on this technology, which is very encouraging. And our message is we can be a great infrastructure platform partner. Our message is we can provide a new payment and settlement layer that enhances the value proposition for your customers. And this is important because these institutions maintain relationships with households and they maintain relationships with corporations, and they need to provide the best service possible to those customers. And if part of that is adding new capabilities that allow those customers to do new things or to do things more cost effectively, that's a benefit to those customers. And the forward-thinking financial institutions are going to embrace the fact that, hey, over here, we might make less fees, but we're going to retain our customers. And our core business of issuing credit or selling bonds or providing other forms of working capital or providing other transactional services or providing investment banking or all these things, these remain key. And in fact, in each of these areas, there's opportunities to enhance those with digital assets. JPMorgan issued a commercial paper digital token bond and sold it -- it was entirely on chain. They sold it to the purchasers who purchased it with USDC. The proceeds of the USDC went to the issuer in USDC. It was an end-to-end USDC transaction with a digital token bond. It was like $100 million. Now the point is that all of this -- all the business practices can move to use this technology. And this is not like an existential thing. This is a -- if I was Fox or Time Warner or AT&T or Verizon or I was Oracle, and I was facing a rearchitecture on the Internet and I can either move fast or I can move slow or I can wait, and that will determine your ability to adapt to the new product delivery and unit economics of this new system.

Really exciting, Jeremy. We started the discussion over and talk to stablecoins. The discussion shifted to a much bigger and broader set of topics and really interesting to see this, as you would say, this huge surface area emerging at the intersection of a lot of things, blockchain, agentic AI, tokenization, really exciting times. And we really, really appreciate you being with us today. It was a great discussion, and I definitely learned a lot. So thank you very much.

Really my pleasure. Thanks for having me.

Thank you.