Intercontinental Exchange, Inc. ($ICE)

We have Stuart Williams. Stuart is the Chief Operating Officer of ICE, the company behind the New York Stock Exchange, world's leading energy futures complex and a growing mortgage technology business. Lately, ICE has become one of the more aggressive traditional players moving into more innovative asset classes like prediction markets and digital assets through its investments in Polymarket and OKX as well as its own plans to build infrastructure for tokenized security. So Stuart, thanks so much for joining us.

Appreciate the invite. It good to be here, Patrick. Yes, sure. So good morning, everybody. Good to be here. So good morning, everybody. Good to be here. Look, cybersecurity has been a key part of our infrastructure right from the outset when we started the company. And so it's been an area that we've invested in heavily because -- as you will imagine, given the assets that form part of our group, we are a natural target for cyber attacks. And so this is an area we've had a lot of innovation and a lot of investment in over the years. We've been leveraging AI ourselves, both in terms of our red team activities and leveraging models to attack our own infrastructure to identify vulnerabilities -- but once Mythos was public, we approached Anthropic, and we've been accepted into the glass wing program. So yes, right now, the team is busy integrating methods into our cyber tools so that we can start looking for vulnerabilities right across our platform. And we'll then start to respond to that as we go along. But it is an important area for us, and so we're pleased to be a part of that program.

All right. Great. So I guess diving in from your operational seat, as we look across ICE's business and the momentum you're seeing today, how have some of the macro factors, geopolitical uncertainty, a new Fed leadership, capital markets activity, how has that kind of shaped the activity and the conversations you're having with clients across the platform?

Yes. So look, we've just announced or recently announced the best quarter in our company's history. And I think one of the things that's really encouraging for us is the breadth of the sort of contribution to those records. So record energy volumes, record interest rate volumes record clients on our mortgage technology platform. In fact, our mortgage technology platform had a record quarter that since Q4 of 2022. And so when you see every component and record demand for our data center capacity, so when you see every component of the business responding and contributing to a record quarter, that's to your point about some of the things that are impacting the near-term drivers and we'll get into some of the long-term drivers a little bit later, I'm sure seeing of the energy supply chain as a result of that is driving a lot of volume in our energy markets. We went into the year expecting a number of rate cuts across all major currencies, and the market is now reconfiguring around new expectations of rate hikes later in the year. So we're seeing and we're also seeing a diverging set of expectations across the ECB and the Fed and the Bank of England and all of that is playing out in our markets. We're seeing AI, as I said earlier, driving increased demand for our proprietary data and for our data center capacity, and then on the mortgage technology side, we're seeing, again, continued demand for greater efficiency, leveraging of AI as we roll that out. So really right across the business, we're seeing good growth.

Yes. So let's just double-click there on the data center piece you mentioned. In Jeff's shareholder letter, your CEO shareholder letter earlier this year, you mentioned that you've doubled your data center capacity since 2020. As we see more AI-driven demand for low latency data access, how do you think about the co-location business as a stand-alone revenue opportunity? And how does that fit into the broader data and infrastructure strategy for ICE?

Yes, it's a great question. Just backing up a bit. I think as many of you will be aware, our really strategic view of infrastructure was that, that was always something we wanted to control and own our own destiny when it comes to infrastructure. So we, for a long time, have really lend into building a high availability, low latency network that really is the ecosystem around our core data centers. And so if you think about every new market that we launch every new customer that joins ICE, every new data element that we generate, all of that increases the value of that data center ecosystem. And our customers have consistently told us that they prefer the ICE ecosystem in terms of high availability, low latency, on-prem infrastructure as a place for them to come and interact with our exchanges. And to your point, we've doubled our [ colo ] capacity since 2020. And we've sold out on Hall 5, which is the most recent hole to go out. Hall, we're busy selling at the moment. That will be complete by the early part of next year. And we've just completed the topping ceremony which is an interesting savory when you finish the building on a second data center building next door to our existing data center building in the Malwa campus, that's going to further provide a doubling of the capacity we have today. And then we're also about to announce a new [ colo ] facility in Chicago for our derivatives business. So we're really seeing our customers join us on this journey. And obviously, they are becoming more hungry for more data. They're becoming more hungry for more computes, and of course, they want to continue to be proximate to our matching engines. So it really is providing that ecosystem. But to kind of wrap it all up, the real benefit of the colo business is that it becomes a multiplying effect on everything else we do because, as I say, it creates that network and every new data element, every new market we launch, the customers have got access to that from their existing location.

Sure. So you mentioned the scaling that you've done. How do you think about the appropriate level of data center footprint for the business? Like is this something where you feel like with these various locations, you've you have enough room to satisfy the demand that you're seeing right now? Or is it something where in 5 years when we may reassess this and you think that it's going to be even larger.

One of the first things we did as we started to see the increased demand in this part of our business was introduce the concept of forward auctions. We're a markets company. You got a commodity that is in short supply. Well, you create a market for it. So we've done that, which gives us good visibility of the demand our customers have for colo capacity 2, 3 years out. And so we're going to continue to monitor that. Right now, we feel pretty good about the investments we're making that we'll both be able to fulfill the revenue opportunity that those investments will unlock, but also we're keeping pace growth as our customers demand grows. So we'll keep an eye on that, but we've got good sight of the demand curve now going forward.

Okay. Switching gears back to the exchange business. I want to touch on energy. It remains one of the great open interest stories. It's up 8% year-to-date already. There's a lot going on in energy markets, as you mentioned, how confident are you in the sustainability of the growth we've seen so far? And then as we look ahead, what is the next leg of the energy story for ICE in your opinion?

Yes, it's a great question. You've heard Jeff say sometimes that he felt the need years ago to diversify the business so that we weren't just dependent on energy and yet that is still our biggest grower. It is -- so I think an important place to start here is we -- the thesis for our energy business was never that we would just move from 1 geopolitical crisis to the next. It was really around building a product ecosystem that would have liquid benchmarks at all the key intersections between geography and energy type and then roll out around those benchmark contracts, a series of differentials and spreads and regional contracts that would allow participants to hedge both long-term risk as well as the more precise needs of near-term risk around particular geographies. And so that's the ecosystem that we've built. And really, that thesis continues to play out. With every geopolitical crisis, we see a rewiring of some of the energy supply chains. When Russia invaded Ukraine, you saw within a year to 18 months, Europe significantly increasing the LNG regasification capacity because it needed to become less reliant on Russian pipe gas. We're now seeing right now with the closures of the Straits of Hormuz, a lot of investment going in the UAE is building a new pipe that will come around the states of hooves, go straight to [ Fujairah. ] The Saudis are building additional pipe capacity to go out to the Red Sea. So every time there's a geopolitical crisis, more resilience gets built into the energy ecosystem. What that means is a practical matter or from a market perspective is that ultimately, it's more hydrocarbons in transit over more trade routes more of the time. And all of that ends up being hedged through the markets that we've launched. And so that long-term tail effect, if you like, of geopolitical impact is that those reconfigured supply chains will add additional capacity and demand for our markets. Now on top of that, there is also the underpinning broad economic growth story, which over the next 20 to 30 years, most economists will forecast. We need to just about double energy. Just to meet the demand curve for both economic growth in Southeast Asia, but also the demand growth required for AI and data center capacity. So all said, we see a strong growth in demand for the next 20, 30 years for our energy markets. And I also think, importantly, there's been a real -- more realistic conversation around more energy from every type is needed in order to meet the forward demand curve that the world has for more energy. So all of that to say is that we're pretty confident and happy with where our energy markets are right now. We'll continue to innovate around that same philosophy of looking for any new intersections between energy type and geography and making sure we can build a liquid benchmark in that location.

All right. So sticking with the topic of energy. Earlier this year, we saw the Iran War breakout. There were people who are looking for access to energy exposure on the weekend, which is when the war broke out and these perpetual futures markets volumes kind of explode. They're still very small in relation to yourselves and CME, but it's something that shined a light on demand for 24/7 trading in energy. Your closest futures peer has kind of took a hardline stance on their last earnings call, saying that perpetual futures were something that they didn't believe shouldn't be traded in the U.S. They were more of a speculative product ICE has been more forward looking, I think you announced with your partnership with OPX, that you're going to launch perpetuals on crude oil, and those would be for overseas customers. But I would love to get your take on not only the opportunity that 24/7 energy trading could provide for ICE, but also just in general, what went into that decision to launch purpose with OKX general thoughts on whether we see perps and energy markets in the U.S.

Let's start with the purpose question, and then I'll answer the 24/7 question second. So I think the jury is still very much out on what the economic purpose for Perpetual Futures is. There's no question that there's a retail desire for near-term speculative an ability to speculate in the near term on what's happening in Iran and with the Strait of Hormuz. But whether that will continue once the energy market settle down is obviously to be determined. What I will tell you is, as I referenced earlier, we spent a lot of time with customers on product development. and looking at what new products they need. So we talk about Brent as if it's one contract. And when we talk volumes, we are talking about the main brand futures contract. But I'll tell you that there are 400 million other Brent derivative contracts that we've got in our complex that will average over the month, that will have a balance of month that will provide spreads between Brent and other contracts. So every time our customers identify a real economic need or an economic exposure that's not being fulfilled through the current contract that we launch another contract. And so if the discussion around perpetuals gets to the stage where an institutional trader or a commercial customer says to us, hey, there's an economic reason to have that contract that isn't fulfilled with the existing contracts. We'll be best placed to launch that, and we could do that tomorrow. The infrastructure is in place, we could do that. The question is whether and what problem we're trying to solve beyond that near-term retail speculative desire. Now all of that said, as a retail product, what we're excited about in the partnership with [ OKEx ] is that there will be products of this nature that are interesting to retail traders that a commercial trader would never touch. But if that scales, we want the ability to participate in that and to get exposure to that. So the partnership with OKX and what we've announced with OKX around the licensing of Brent and WTI for retail-based pops on the OKX platform is a start in that direction. We'll learn a lot from that. We'll learn once the war ends, whether there is actually long-term desire for a contract like this, whether the attention goes somewhere else but we'll have that optionality. And then the last comment I'd say is on the point of bringing perpetual futures or purposes as an asset class or as a contract type into the U.S. It is going to require some changes to the way these platforms operate and the convergence of innovation with regulation is something that we're seeing play out at the moment, and that has really been a sweet spot for us. If you look back in history, when the CDS market hit the financial crisis and something needed to be done to help unpick the complexity of that market and create clearing, we provided that. We're now 95-plus percent of that market. When LIBOR needed to be reconfigured and ultimately replaced, we enter that market and help facilitate that so really when innovative markets start to grow up and start to come into the regulated environment, that really is a sweet spot for us. And which is also part of the reason we've partnered up with OKS.

In the U.S., do you feel like this decision by the CFTC to grant the Bitcoin perpetual future for cal sheet? Do you think stops at crypto. I think that there is going to be pushback from a lot of people in the industry if you start putting perpetuals on real markets, if I could use that term. Is that something that concerns you from an operational standpoint? Or what have you had any talks with regulators about this?

Yes, we're certainly very active in discussions with the CFTC and the FCA and the MS and you pick ultimate soup. We're speaking to all of the regulators about these markets. again, if there is an economic purpose, if there is an economic problem, we can solve with that contract, I feel confident that there's a path for them to be regulated and introduced. If there's not an economic purpose, then we probably don't end up going down that path. Obviously, we'll learn a lot through the [ Kelsy] contracts as they get rolled out in the U.S. and conversations will continue. I do think, though, for commercial participants, the economic structure of a perp, which really reflects the daily spot price or in energy markets actually have the derivative on a derivative. It looks very much like an ETF actually in energy markets. Doesn't fulfill the need of a commercial hedger to understand the forward curve and on the basis of a forward curve make decisions on investment. So what customers are doing today is they're looking at what the Brent price is 5 years from now in order to make an investment decision on a particular field or a particular refinery. If I'm just determining whether I'm going to move a cargo of oil from the U.S. Gulf Coast to Asia, I'm looking at the forward curve to determine what the price in Asia is going to be in a month's time. That's the journey time versus Europe but perpetual market doesn't give you that. It is a daily spot market. It is not a forward curve. So it really -- that's why I say it's very difficult to see how a perpetual market creates anything that's useful for our current customer base. For retail speculating on the outcome of war, sure. But for a commercial customer who is making forward investments based on our current view of those prices, it just doesn't fulfill that need. So it's a different use cap. Sure. No, I would definitely agree with you. I think it's more of a retail product than maybe some it's going to fill the role of a tool for speculative retail for the most part. And it seems like that's what it's been so far.

Okay. So moving on, I want to double click though on Polymarket and OKX. how do you think about leveraging those 2 relationships kind of across the board outside of person things like retail distribution, data, market structure? And where do you see the most meaningful commercial overlap with those 2 partners?

So I think the commercial. Initially, the commercial overlap between those 2 is, if you think about ICE business, we've been historically a B2B company. We've got a fantastic commercial network, fantastic institutional network, we don't have a retail distribution network. Both of these platforms have got great retail distribution. OKX has got access to $120 million crypto-native retail customers predominantly in Asia. That's not a network we have today. So in addition to that, though, what Poly brings is the whole notion of crowdsourcing a view on a particular outcome. I'm not talking about sports betting. That's not the part of the business that we're particularly interested in. The part of the business we're interested is the non-sports betting side, the production market on particular economic related events. That's, we think, is an interesting concept. We think the idea of having markets on all manner of economic indicator is an interesting idea and could have real economic utility. And so from the poly perspective, taking that data as the only distributor to a commercial audience of that data that we think is very interesting. So what we're doing right now is we're taking the polymarket data, and we're normalizing that, structuring it and pushing it through our consolidated feed our commercial customers, who, for years, have looked for alternative data sources to inform trading decisions, and this is yet another alternative data source that is of interest to that cohort. The other thing on the OKX side and of course, as we've said -- sorry, you're going back to Poly. Of course, on the polyside, they've got a pure crypto-native platform execution through settlement that manages everything as part of those smart contracts, which we're learning a lot about along the way. On the OKX side, we talked about the retail distribution that we've got there. And then, of course, the other interesting thing that OKX brings in addition to the retail distribution is the -- again, the crypto-native settlement capability that they've got which, of course, we're looking at from a 24/7 perspective on the NYC to stitch together Pillar, which is our execution platform with a blockchain native settlement layer that can facilitate 24/7 trading at NYC.

Okay. I want to shift gears you took an early lead with extended average trading in equities on Arca alongside the DTCC. Where does the industry sit today on a path to genuine 24/7 trading in equities and what are the operational lift items that ICE still needs to get solved to make that a reality?

Yes, it's a great question. So we -- obviously, we took the lead on ACA and that has extended hours already. In our derivative businesses, we already have some spot energy markets in Europe that operate on a 24/7 basis. So we already have 24/7 platforms with the ratio is my point. Where we are now is engaging with our commercial customer base on where there's going to be interest in a kind of longer window for trading. So we're already open on a Sunday evening. And so we're in discussions with the market right now on is there benefit in us opening Sunday, we've been earlier. Do we open Friday a little bit later. Where are the -- what are the pros and cons of those extended hours. And of course, that would require the market to also have operational staff available to support trading in those hours. So that conversation is ongoing. But of course, the technology and the operational capability to support 24/7 is there. So I wanted to just make that point. I think the more interesting innovation, though, if we circle back around to the NYC, we've got extended hours on Arca. My point around taking the Pillar platform, which is a platform that is able to and today does handle north of 1 trillion messages per day at sort of a 10-microsecond latency and stitching that platform to a blockchain native settlement layer which is what will ultimately facilitate true 24/7 notional-based trading all around the world. of the NYC equities. And so that is an area of a lot of innovation. There are things we still have to do. There are some regulatory approvals we still have to get. There are some APIs and connectivity we still have to put in place, but all of that is manageable, and we're looking to complete that in the coming months.

All right. So we'll end on AI. It's transforming workflows across financial services. How is it showing up in ICE's business across mortgage Feds and exchange business? And where do you see the biggest opportunity, I guess, to deploy it across various businesses?

Right. Great question. So let me start with the infrastructure. I talked a little bit earlier about our data centers. So we've acquired a bunch of GPUs, and we're in the process of racking and stacking those within our data center. So what that's going to give us the ability to run AI on-prem. We obviously have AI -- we have ability to run AI in the cloud, and we'll be able to move AI workloads between the cloud and on-prem. So we can optimize the speed to market and we can optimize for cost. That gives us some great levers in terms of ensuring that we can control the cost of running AI. In terms of the internal use cases at ICE, we've really centralized all of our AI capabilities under IS Aurora that you've heard us talk about before. So we are using AI right now for coding use cases. All of our developers are using AI every day to write code. And that's definitely one of the big reasons why we've been able to accelerate some of our move off the mainframe and some of our synergy execution on the post Black Knight integration work. On the product side, we've got MCP servers that we've launched in Cloud, and we're in the process of launching MCP servers as a channel. The way we think about AI from a fit perspective is really as a channel. It's another distribution channel to get data to our customers and another way that customers can have the ability for AI to directly access the data and process that data within their workflows. So that is an area of innovation for us right now. MCP service hosted agents that we're busy rolling out across the Fitz plant. We're starting out with the nonproprietary data so that we can learn more about some of the controls that can be placed around that data. but the idea is over time, that will extend to a full set of data. And then on the mortgage side of the house, this is where we're seeing some real opportunities. And let me give you some examples. So we've launched 16 automated workflow exception-based workflow AI agents. And what I mean by that is, if you think about the mortgage workflow, which is incredibly manual-intensive -- the big opportunity for AI there is taking a lot of those manual workflows and translating them into exception-based workflows where the human just interact at the point at which a human input is required. So there's an exception as human being needs to go and have a look, check something, sign off and move on. And so we've already got 16 of those agents operating within servicing. And if I take one of those. So one of the workflows, which has 46 touch points that normally takes 10 days is now down to 6 touch points over 2 days. That's real efficiency. And so we're continuing to roll out more of that, and we've got -- we've just launched in March, chats and voice AI agents to support the customer support part of that business. And so we see this as a real opportunity. One thing I would say, though, on AI in mortgage, mortgage is one of the most regulated industries in the U.S. right? This is retail making one of the -- maybe the biggest investment they will make in their lives. And so there's a lot of concern around that. So 2 important things when you think about a heavily regulated industry. The first is that there are like 1 million pages of regulations across federal regulation, state regulations; county regulations that need to be understood and triage to make every decision. So that is a perfect use case for -- but at the same time, there's a plethora of requirements and new requirements that are coming out a little time for how AI is used. So the responsible rollout of AI is critical. And so we, again, are taking a very thoughtful approach in how we roll out AI. We're making sure that within those mortgage workflows there is clear auditability. There is clear observability of all the decisions that are made along the way and so that our customers can use that both to gain the efficiencies but also have the confidence that it's being done in a way that's secure. So there'll be more to come on the mortgage side, but that is an area that we're already seeing a lot of gains.

All right. Well, I think that's all the time we have. Stuart, thank you so much for joining us. It's been great.