Okay, wonderful. Thanks, everybody. So we'll get started with our next presenter. It's my pleasure to welcome back Jeff Sprecher, Chair and CEO of Intercontinental Exchange. Over the years, ICE has evolved into a leading global network across trading, data, and mortgage services, with lots of technological investment in the space. We look forward to getting Jeff's perspective on the evolution of today's markets, growth opportunities for ICE over the next couple of years, and of course, just broader thoughts on the environment. So Jeff, thank you for being here. This company's been around for 36 years. I think you've been here for at least the last 20 years.
I've been around for a lot of them.
So I appreciate your time and the support you've given us over the years.
I always feel like, you know, the guy has acquired some knowledge, so.
Hopefully. Well, we need it today probably more than ever. So, look, my first question is actually a bit related to that, which is, you know, broadly around market structure. You know, you have a long history of building traded markets, really starting with the electrification of global energy and expanding then to fixed income data and obviously did try to digitize the mortgage industry. We might be potentially again on the cusp of another fairly big shift in the space with respect to blockchain technologies and tokenization of assets. So it's a bit of a big picture question for you first, but how do you expect the ecosystem to evolve over the next three to five years when it's real, you know, related to tokenization of traded assets? What, what's real, what's not, who wins, who gets disintermediated?
Just bigger picture thoughts from you on this topic would be super helpful.
Sure. Well, first of all, thanks for having me, year after year. I appreciate it. I think the biggest. I do think that tokenization is gonna have a big impact on financial services. And I think its use case that will emerge and be dominant is essentially movement of capital. I have because I've been around a long time. I actually have access to a lot of important and powerful people that are in the U.S. and the U.K. and the E.U., Middle East, and Asia. And the last year or so, I've talked to a lot of regulators and central bankers and people that run infrastructure or that manage and regulate infrastructure about tokenization. And sort of nobody that's running the legacy systems for governments wanna touch it.
So as a result of that, you have all these entrepreneurs that are building tokenized systems and now government catching up to try to organize and regulate them. But I think those cats have run out of the room. I say that because, you know, I've suggested to some regulators that if the Fedwire, if the SWIFT system, if the ACH systems worked outside of banking hours, then there would be no reason to have tokenized money. We already have digital money, but and, you know, a young person today will probably never carry paper or coin-based fiat money. But the banking system closes, you know, during bankers' hours. I think what tokenization is going to do is allow 24 by 7 by 365 movement of capital.
And eventually then it's gonna be figured out how to be regulated. When I ask bankers, central bankers, you know, why wouldn't you do this? The answer is, you know, the infrastructure that we manage is so rigid in the way we start it and stop it. And, oh, by the way, if there's a Silicon Valley Bank collapse, we need the systems to be closed so we can resolve it. And, oh, Jeff, you do a lot of M&A and don't you need the weekend to do that so you can have a surprise Monday? And, you know, that's kind of baked into regulation. But meanwhile, while trade is being rebalanced, physical trade is being rebalanced, capital is probably going around the world at a velocity like we've never seen. And the world is becoming more dollar denominated.
And with Tether and Circle and other stablecoins that are being invented, you're sending dollars to foreign places and it's becoming. I think Tether may have as many as 500 million users already that are essentially doing commerce in U.S. dollars in small countries in rural parts of Asia. And I think that trend's gonna continue.
So that's much more—so you're talking about more of that as a concept of, you know, money.
Yes.
When you think about it in the concept of trading assets, securities, equities, options, futures, and the effect that that might have on the ecosystem as well.
Yes.
So curious, I guess. Wonderful on your thought there. And then, secondly, what role do market operators like yourself will play if there's a wider trend towards tokenizing things other than cash for all intents and purposes?
Yeah. I mean, you know, the basic tenet of trading and we have 13 exchanges, including the New York Stock Exchange. The basic tenet of trading is that buyers are looking for sellers and sellers are looking for buyers. And so the Romans created a town square where everybody gathered. The New York Stock Exchange started underneath a Buttonwood tree on the corner of Wall and Broad. People gather in one place. There is this concept of let's go DeFi where everybody's everywhere and we'll have a browser that will search the universe and find you. And that can work for kind of certain assets, but the reality is there's a latency in searching around the world for everything.
And generally speaking, history has shown that buyers and sellers would prefer to gather in one spot so that they're all together and that things work on a FIFO basis so that no one has a timing disadvantage. And so I think, you know, the legacy—so the New York Stock Exchange, for example, which started in 1792, you know, has seen the invention of electricity and, you know, the smoke signal and eventually computers, big computers, small computers, handheld computers, high-frequency traders, people bouncing things off the ionosphere in order to have an advantage. And so it's incumbent on us as the steward of the New York Stock Exchange to adopt the latest technologies, so that people have access. But I still think that there's something about gathering.
It doesn't mean new exchanges won't be formed, but I think it's incumbent on legacy operators like ourselves to be mindful and be at the forefront of adopting technology.
Yeah. Speaking of new technologies and maybe new markets, let's talk a little bit about Polymarket and the investment you guys made there recently. $2 billion. It's the largest non-sports prediction market in the world. First, I guess, talk to us a little bit about the strategic rationale behind this and what your long-term aspirations from this relationship looks like.
Yeah. It's exactly what I just said, which is we studied all of the tokenized and DeFi companies. We did a pretty deep dive. We have a lot of conversations in the background. We're very fluent in who's doing what. And we felt that Polymarket was really the leader in DeFi protocols of anybody we'd seen. And we wanted access to their engineers. We wanted to get educated on exactly how they were doing what they were doing. And then it turns out that, you know, they needed to get access to the United States, regulatory access to the U.S. And in order to be compliant in the U.S., they needed to do more of what we were doing.
And so, you know, we shook hands and said, "We'll help you, you help us." And so it's been really, and this is all about technologies. This isn't about sports betting or predictions or anything. It—this was about let's get fundamental technology knowledge transfer going on. And the thing about, Polymarket does two things that we don't do. One is, let's just take a sports bet. There can be a person watching a game and they can say, "I think this person's gonna miss the next shot." And I'll bet $5 on that. And so, they can post that and it can be disseminated broadly and it will also be discussed in a social media setting so that everybody knows, you know, Jeff thinks this guy's gonna miss the shot.
And then somebody can take the other side of that. You can see the shot, somebody gets paid. And that all happened, that contract could exist for an hour. That is not unlike anything that exists on traditional exchanges. There's no such thing as something that comes, exists for an hour and disappears. And in fact, in the U.S., the Commodity Futures Trading Commission requires that we give them 30 days' notice before we list anything. I was told that over the U.S. government shutdown, there were roughly two million contracts that were created because of this sports betting and prediction markets. And we had six contracts that were at the CFTC that the CFTC put a hold on and didn't approve. And they were grades of oil and natural gas that were going to be delivered in various locations. Okay?
So our regulator is dealing with that. And you have this other trend going on of short-term, you know, creation before the regulator could even act on it. And so one thing that we wanted to get with Polymarket is, how are you doing that? How are you disseminating that? What does your regulatory conversation look like? And can we help each other? Because the market wants more dynamic creation of tradable products. The other thing, which is the more important thing and what I alluded to in your first question, is that once the shot is made and buyer and seller exchange value, money moves. And it moves via a smart contract. So there's no human intervention. And it moves via a token that represents cash. So is accepted by both parties as being worth a dollar.
That is really novel. I mean, that is outside of a clearing infrastructure, outside of a banking infrastructure, value just transferred and title transferred with a bearer instrument, that is unregulated and essentially represents tokenized deposits. So that's coming, I think. It exists. And I think it's gonna be institutionalized. But what it's gonna unlock is 24 by 7 by 365 capital movement. It'll make, like this audience that are largely institutional investors, it'll allow you to not keep excess capital in various jurisdictions where your companies do business. Somebody like a Goldman Sachs that is almost omnipresent in every jurisdiction, and has local subsidiaries, the local balance sheets and what have you, that your treasurer has to keep capital all around the world for the various businesses that you do and the risks that might occur during daylight hours.
You're gonna be able now to, you know, keep capital maybe in a central treasury and move it on demand at a moment's notice using the internet.
Yeah. No, it's a fascinating topic. If you think about commercializing that relationship for ICE beyond the value of Polymarket, which, presumably you guys already actually made some paper money on, just based on some of the press announcements. But if you think about the data.
Turns out people like our, you know, like our pedigree actually helped those guys, you know? So, you know, for all of us that are in legacy finance, like we actually, you know, have a second life.
Yeah. Self-affirming. I'm intrigued by the data that comes out of this marketplace and how you guys might use it. And then when you talk about the new markets and creation new markets and perhaps institutionalizing that, what are you thinking about in the intermediate term, right? Something that's tangible in the next kind of two years.
Yeah. So one of the fascinating things that Shayne Coplan and his team built is, imagine there are tens of thousands of event contracts that exist now. If you've got users that are saying, you know, will somebody invade Syria? Will there be a shipwreck? Will, you know, all kinds of things that people have an interest in that are willing to put their capital behind. And one of the things that Polymarket has done is they have an AI algorithm that is just looking over all of that content and finding something that basically a market that's gotten hot where there's a lot of money coming into it or a lot of conversation on social media. And it will put that market at the top of a list.
And so if you have their app or if you're on their website, you're constantly seeing new markets pop up to the top of the list that have a lot of attention to them. What that does is, it's hot news. It's it takes the kind of news that you're used to seeing on Twitter or X and then imagine that the wisdom of crowds are speculating on the value of that news and putting it at the top. My understanding is that when the U.S. bombed the nuclear facilities in Iran, that in the war room, they had Polymarket up with the odds that the U.S. would bomb Iran. Why would they do that? They're not betting. They launched a mission, but it took 20 hours for those planes to get over there.
And they figured that if somebody could see it or figure out that there was lots of pizzas being ordered at the Pentagon or whatever, that they would start to see activity there and it would give them an indication that maybe, maybe somebody's figured it out or news has leaked or something. And so what we're seeing now, to answer your question, is hedge funds. Hedge funds have the Polymarket app up and traders are using it for hot news. We think for our kind of legacy business of oil and gas and cocoa and coffee and, you know, all kinds of basic commodities, even interest rates, that people are gonna wanna see, okay, there's something hot here and I better pay attention to it that can impact the risk on the position that I'm that I either have or about to take.
We've got probably over 5,000 institutional customers that we think would want that. We have. We probably serve 10,000. We think 5,000 of them or so would be interested in this. We're helping to take this data, which is unstructured, you know, real time. This market got created real time. There's activity real time. It was deemed to be worthy of dissemination. We're working with them to how do you put that into an institutional environment that's used to just, you know, receiving what is the price of the S&P 500?
Yeah. It's fascinating. Look, maybe bringing it back to some of the traditional markets, let's talk, let's talk about your energy business for a couple of minutes. You know, despite the fact that it's, I don't wanna call it the most mature business because it's certainly not really growing like a mature business, but it's your kind of original business, but the growth really continues to be very phenomenal, right? Like 2015, I think revenue's up another 15%-20%, something like that. And importantly, OI is still growing really nicely. We talked about a lot of structural underpinnings to that growth, so maybe as you look out into 2026 and 2027, what do you think is accelerating structurally? What's decelerating? How would you guys think about growth in that business over the medium term?
Yeah. It's amazing how that franchise, like, I partly pushed the company to diversify because I thought, well, how many people are gonna trade oil? I mean, once you have all the oil companies and some of the sovereigns that are there, maybe some airlines, like who else would ever trade it? And it just continues to grow and grow and grow. And part of the reason is the answer to your question, which we've seen the center of gravity of energy risk management move east. You know, I don't know, 20 years ago it was about U.S. oil and gas.
And then, you know, we made a big bet that gas was gonna globalize, liquefied natural gas would be put on tankers, would create a common market around the world, which is what's happened, largely also driven by Europe's energy need and kind of exacerbated by cutting them off from Russian energy. And it continues to move east into Asia, which is growing. Part of it is population growth and just, you know, the energy dynamics of a population that is growing in certain regions faster than the West. But also I think a lot of these trade rebalancing deals that the U.S. is striking, whether they're trying to get equal transfer payments is, hey, we'll take your goods, but you've gotta take our energy.
And there just seems to be a more robust intercontinental movement of energy at a moment in time when everybody's building data centers, you know, we're all becoming more electronified with everything we do, you know, natural gas is the transition fuel away from coal and oil. And it just seems like it's been, you know, continued limitless growth. Our, I think our open interest in our European natural gas contract is up 46% year over year versus the year before and the year before and the year before where it's been double digit, double digit, double digit. It's just almost hard to fathom. But it is this, I don't know, retrenching of trade coupled with a movement of growth and infrastructure investment that continues to move east.
Got it. Great. Well, it sounds like lots of runway there still. Look, I'm gonna go around the ICE's ecosystem, a little bit here 'cause there's a lot to cover. First, I wanted to start with a question on data centers. It's not something that we spend a ton of time talking about as a revenue driver for ICE, but it has become one kind of maybe perhaps quietly so. I know you guys build out extra capacity. Talk to us a little bit about monetization strategy for your data center capacity. How big is it? Where do you see that going over the next couple of years and who are the customers?
Sure. So going back to the, you know, the Romans creating a town square where everybody got together, these exchange data centers are, you know, where everybody is actually physically located. The New York Stock Exchange trading floor is not, is not the epicenter for stocks anymore. It's the data center we have in New Jersey. And so we were fortunate in a way when we bought the New York Stock Exchange. They had decided to build this really beautiful big data center in New Jersey in an area that had access to electric power. And we started to offer. We stripped everything out of that other than the trading infrastructure that we run at ICE. And then we've allowed third parties to co-locate there.
So you have high-frequency traders, of course, but you also have brokerages, Goldman Sachs, anybody that's offering brokerage, wants to have, you know, their infrastructure close to the actual match of the buyer and the seller. And now you've got with AI this demand for more decision making that's making its way into those data centers that's closer to the match. I met yesterday actually with one of the most prominent algorithmic trading firms in New York and asked them about, you know, what are they doing. And they're putting so far that they're not allowing AI to actually make big risk decisions. Those are still being made by humans back at the office.
But they are taking the data and organizing it on- prem in the data center and sending more results back as opposed to the raw data. And secondly, certain low-level kinds of ordinary trading decisions they can make that they're willing to put, you know, their risk management into that algorithm. I think that's like early days, you know? I think you're gonna see more of this as people get more comfortable with these models and allowing them to actually put capital at risk. It'll take some time, but I think it's an inevitable trend. So we've decided to build and announce we're building a second data center. The way these data centers work is you build this big building and then we call them halls. I think we have seven or eight halls.
And anyway, you know, most of the building is just an empty concrete slab, but you start building out one little area, and then selling rack space in that area, which includes power and cooling. And then organically you kind of go, you know, build the next hall, the next hall, the next hall. And that allows you to innovate 'cause there've been lots of trends in cooling, lots of trends in energy management around data centers. And so it keeps you kind of at the cutting edge with each new investment that you make. And so we'll probably have a 10-year runway with the data center that we're building now. We auction off space once a year. We literally have an auction for who does it. It's very transparent.
So people, I don't know, they kind of like the transparency and the conversation that we're having about what space do we have and how do you get access to it. Tends to put the highest and best use in those data centers by doing it that way.
Gotcha. Great. Let's talk about mortgage for a couple of minutes. You guys made an important bet there that is, you know, has been paying off to some extent, but there's obviously been some cyclical headwinds.
Yeah.
They've been pressuring growth there as well. It, it feels like we're coming on the other end of the cyclical dynamics, hopefully with both lower rates and you already started to see a pickup in refi activity. Any thoughts about what that means for mortgage revenue growth for you guys into 2026? And then more importantly, when it comes to new product development and new initiatives, how do you sort of think about the longer term organic growth algorithm in that business?
Yeah. So, you know, we've built this network essentially and then by stitching together a bunch of legacy businesses that we acquired and moved the whole thing onto what I call a mobile phone plan where you sort of buy a tier of volume that you can use it for. And if you go above your tier, you owe us a minimum, in any event. And if you go above the maximum, then there's a per unit charge. And that becomes a high per unit charge. So it was funny that we started moving people into the cell phone plan a couple of years ago and most of the lenders we have about 3,500 lenders now that are on this.
Most of the lenders went to a higher plan than their last year's business thinking, oh, you know, we're good. We're gonna grow. We're going onto this network that's gonna help us take costs out. And so we'll go up a tier. With rates staying, you know, higher for longer, now we've seen people go down a tier. But what they've done is they've gone down a tier in fixed price, but they're gonna pay a higher per unit price when volume starts to come. So the thing is spring loaded for kind of a, you know, a surge in housing.
We look at just historical trends, you know, and we really believe things will revert to a mean that, you know, a pre-COVID, any pre-COVID year that you would pick, 2019 or what have you. It's fundamentally different than what's going on in the space today. I happened to have dinner last night with somebody who's a very prominent, well-known person in the mortgage space. I asked him, like, what do you think the catalyst is? Their economists think that mortgage rates have to have a five in front of it, that there'll be a psychological release when mortgage rates start with a five. That's their guess. It's not based on any knowledge other than anecdotal from talking to thousands of clients.
Whether we get to a five is hard to know. Obviously, the Trump administration is pushing hard on short-term interest rates and whether how that transpires into mortgage rates, we'll see.
Yeah.
But I don't know. We feel good about the business, you know? We just not, but not so much about the environment.
Are there any areas of potential disruption risk you see in the mortgage franchise? So, you know, we talked about tokenization, obviously.
Yeah.
The concept of blockchain is super relevant for mortgage, you know, just thinking about how inefficient that system is, and that's part of your thesis as well.
Yeah.
You try to actually digitize it and make it better. So, are there risks to a new technology coming across and disintermediating you guys in one way? And how would you deal with that?
First of all, I think if anyone's gonna disintermediate us, it's gonna be us. We have 3,500 lenders. Every major person you can think of is in some way attached to us. And we're talking to everybody. I think the opportunity, so where there's real opportunity, and where we're getting a lot of conversation is not tokenization, but it's the use of AI models to better predict the home values, the creditworthiness of a client without making the credit decision, which the regulators would not allow in our mind. But more importantly, you know, a lender today tends to loan the money and then sell it to Fannie and Freddie or sell it into the market where it makes its way into mortgage-backed securities of jumbo loans and other things.
So they don't own the loan long term. And that's the norm. And so they lose connectivity with the client. And yet now with models, with AI models, and we're doing this for our clients, we can value the home. We can look at the balance of the mortgage. We can look at, you know, the periodicity of payments that the borrower has and make predictions, tell the original lender, maybe you should contact that person. They'd be a good refi candidate. There's a lot of equity in that house. Maybe you wanna offer them a, you know, sort of some kind of cash-out refi product. Maybe you wanna talk to them. They've been in the house for 12 years. Maybe you should go back and see if they want a remodel loan. Things like that.
Mm-hmm.
That would allow the original lender to have a customer for life even though they sold the loan into the capital markets. I think in terms of tokenization, the regulation is so heavy around mortgages and what you can say and what you can't say and what information has to be gathered. And that's part of what we do is help make that more efficient by digitizing a lot of it. But I do think, you know, this current administration and a lot of lenders are working with us on two kinds of loan. One is, can I sell my house and go to a new house and have my loan be portable? Or secondly, can I sell my house and keep the mortgage that's there and have it be assumable?
Those things are very difficult to do right now because mortgages get sold into capital markets. There's all kinds of rules about who can know, you know, the actual owner of the mortgage that's in a mortgage-backed security and what happens if there's a default and who gets rights. And it's very complicated. It's all designed for consumer protection. But with a token, you could theoretically keep track of everything without necessarily giving up identity. And might allow for these more innovative kinds of loans. Those kinds of loans exist in other countries where they don't have mortgage-backed securities, where the loan stays on the balance sheet of the lender. We're trying to figure out with the industry, can we keep the MBS market, which is a very robust market, but maybe find ways of facilitating some of this other activity.
Great. All right. Fair enough. We talked about AI in a couple of different points in this conversation, mostly related to revenues. But on the last earnings call, maybe it was the one before that, you guys also talked about how you're using AI internally to improve efficiencies within the firm, which I thought was probably worthwhile to spend a couple of minutes on.
Mm-hmm.
So, talk to us a little bit about how you're using AI operationally, what kind of a savings you hope that could produce. And again, more importantly, just contextualize for us, like what does that mean for ICE's longer term expense growth?
Mm-hmm.
Is it just slower pace, more reinvestment capacity? Kind of what does that mean?
Yeah. A few years ago, we set up what we called an AI Center of Excellence, and we hired some really top AI engineers, and so we had a center for domain knowledge. And so we look at every model, and new models come out every day. And so we have a group that in my company is very proficient. And then what we say to all the business people is if you have an idea or a use case, go to those people. They can tell you, you know, what we should be doing, how we should do it, and so on and so forth. What we found as a result of that, we did a deep dive across the whole company. We've been doing this for years now.
What we found is just, I guess, what you kind of expect, which is where there is a language-oriented task, the AI model can help streamline that task. So language being code writing. So we, you know, all of our people now have Copilots, our call centers and help desk and customer support areas. And then some of this mortgage stuff that I'd already talked about, which is really digitizing what was a paper-based application and signature, wet signature process. I think if you mark the market today, I would tell you that our good people are better and more efficient, and that we haven't necessarily eliminated any positions or what have you. We have slowed down just naturally, not organized, but naturally we've sort of slowed down these kind of entry-level jobs.
They have been getting automated in many cases where somebody might have had a junior person doing something, and so I have, you know, I have a lot of friends that have children that are graduating from college with good colleges with good degrees that are having problems and their friends are having problems entering the workplace. I kind of see that 'cause I look at our own behavior, which is, you know, we're kind of slow walking some of that kind of hiring now because we're using these tools. So I think it, for your model, it, Alex, it kind of slows the growth of expenses.
Yeah.
As opposed to being some revolutionized thing. But that's today. I mean, you know, as we know, these models get better every day. And, you know, someday they may be, you may be interviewing one up here.
Gotcha. It will be two models interviewing each other.
Exactly.
To be honest. All right. So before we wrap up, look, I did wanna ask you a question around acquisitions. Understanding we only have like a minute left on the clock, but you guys made an investment at Polymarket. That was one of the more meaningful things I think you've done since like the mortgage venture that you guys have, kind of, you know, pursued. Anything else looks interesting out there? Obviously, I don't expect you to announce anything on stage, but as a framework, and as you think about the types of businesses you guys are in today, how attractive is M&A opportunities for you guys for the next 12-18 months?
Yeah. I don't know. You know, our own stock in our mind is a great value. Like, you know, when you all in the audience here rotated into the Mag 7 and kind of left us behind, it just, you know, we started our share buybacks. We run our own models on our own stock. And if we do M&A, we always test the thesis, the M&A returns against our own share buyback returns. And right now it's been very fruitful for us in our minds to pick up as much as we can. So, fortunately we're generating a lot of free cash flow and able to do that.
Yeah. Great. Nice and sweet.
Great.
Thank you very much.
Thank you all.
All right, Jeff. Thanks. Appreciate it.