All right. Welcome, everyone. This is Andrew Stein, I'm India Business Information Services Analyst here at JP Morgan. This is the Ultimate Services Investor Conference, and this is the Info Services track. We also have a Payment track and a Business Services track. We appreciate the gathering of the best services companies, the best services investors. This is the S&P Global session. What a great time to kick it off. I'm with Martina Cheung. She's CEO now for one year. She's a veteran at S&P. We first hosted you here last year. Welcome back, Martina.
Yeah, thank you. I'm delighted to be here.
One of the big milestones for you is the Enterprise Data Office and the Chief Client Office. My question to you is, is the team as excited about these things as you are? Is this something that's sort of been mandated top down, or is this something that you really see a lot of followership, excitement inside the organization about?
Yeah. Maybe just to start with.
Which I am too. Sorry.
Yeah, yeah. So the rationale for setting them up, which was, you know, at its highest level, ensuring that at S&P Global we can operate with an enterprise mindset. With the Chief Client Office, this is a pretty small group of folks who have a group of around 130 of our most strategic and largest customers that they serve in partnership with the divisions. The idea there is to essentially bring the best of S&P Global and to develop senior C-suite relationships based on trust where we can actually really be very closely aligned with what we're doing for our clients, what we're building for our clients. In the Enterprise Data Office, it is to, you know, there were always two sides to the coin with the Enterprise Data Office.
One is let's bring things together to have opportunities to create more insights with the same data, connected and sliced in different ways, linked for new insights. Let's make sure we can do that with real speed to market. Let's also make sure that we're doing it at the most scaled, efficient way possible, largely because we have, you know, we have thousands of data operators who work with multiple systems. Lots of opportunities to do things more at scale. If you look at how this has, you know, our experiences on these two areas over the past year, perhaps to start with the Enterprise Data Office, we've had a tremendous amount of success on both counts there.
The Enterprise Data Office has partnered with the Market Intelligence team and the Kensho team to bring more data and more insights out more quickly. Everything from in the actual core operations, linking millions of assets together, linking millions of people data and additional third-party data, plus our own proprietary data to get more insights onto our product more quickly. These things would not have been possible if we did not have some of the technology, such as the linking capabilities that we have through the acquisition of Terahelix, for example, as well as the work that Kensho has done with Kensho Link. That has been phenomenal, getting that faster out to market. We have also been able to really make sure that we are making the best of the assets that we have in the sense that enabling that growth without adding headcount.
We said earlier this year that we expect 2025 to essentially be peak organic headcount for the Enterprise Data Office. We expect to be able to continue to grow what the Enterprise Data Office does without growing headcount on an organic basis, of course. That has been quite a success. You will have heard Sagata also say last week that the Market Intelligence division has benefited quite a bit from a margin expansion perspective through some of the work that has already been done in this first year of the Enterprise Data Office. On the Chief Client Office side, the team there has really focused on level setting the relationships across our most senior clients in every sector. With that, we have been able to elevate the dialogue, land some incredibly important overall deals, such as some that we have referenced publicly.
Barclays is an example of that, which is an interesting way of a sort of like a reference point, if you like, for how we are working with many of our largest financial clients. We've also been doing a tremendous amount of work with our large corporate clients, our large energy clients as well. I think with this and with the Enterprise Data Office, where you get the excitement with the Chief Client Office is, you know, every salesperson wants the senior team from S&P Global to come and visit their client and, you know, have that top-to-top conversation because it puts them in a much more strategic position to work with their account parts across the client.
I was telling Andrew last night that when we were in the Middle East a few months back, I had a salesperson walk up to me and say, "Can I get these two clients? Like, what would it take for me to get my two clients into the CCO? I really want to do this." You know, there is a tremendous amount of excitement there. I think with the EDO, there is just massive excitement around, like, the possibilities for technology. I mean, they have just been moving at, like, express speed on integrating new technology, finding new ways to do things more effectively, but to also create new insights for customers.
One of my favorite examples of the power of the Enterprise Data Office is, you know, what we've been able to do most recently with IBM, where it's not just a, you know, an agentic AI story where we have, we're deploying our agents with our data into enterprise systems partnering with IBM, Watsonx Orchestrate, but it's also a data story because those agents are bringing data from our supply chain data, our economics and country risk data, our overall sector data. It is a really neat story that shows the power of the EDO.
You're saying morale around this internally is very good, right?
Very good.
Okay. Gotcha. Okay. Could you just talk about why you're so confident, particularly in MI, where AI is only going to be a source for positive benefits at S&P and there's not really risk because you're mitigating those risks?
Yeah. We, you know, we had said last week that, and we have said, I would say consistently, that we're incredibly excited about the potential for AI. You know, I think a lot of this, I can tell you that the Kensho team were talking to us about generative AI long before, not that I knew what it necessarily was, but they were talking about it a long time before, you know, 2022 when OpenAI came out with ChatGPT.
Generally, I would say the foundational capabilities that we invested in with Kensho over the last many years since the acquisition of Kensho in 2018 and them serving as our sort of AI foundation for the company, those foundational capabilities have proven to actually be extremely helpful in accelerating our own, I would say, journey into the integration of generative AI, both internally as well as into our products. The reasons why we see this as a positive are first, you know, the productivity piece of it, which of course I'm sure all your organizations are looking at as well. It is not just the Spark Assist piece that, you know, we've been talking about. That is incredible in terms of providing access to everybody and getting everybody context to where, as well as, you know, where of how to use the tools.
We have been very focused now on also examining the more scaled and strategic integration of the tools into some of the largest groups of people that we have around the organization where the productivity benefits could be felt most productively. You know, there are places you could probably imagine like technology, EDO, for example, where we are already doing that with some of the AI and agentic tooling that Sagata and team are deploying there. In other areas in the organization, like research, for example, as well. From a growth perspective, look, we are just incredibly excited. We started with making sure that our core products were actually being improved and very rapidly with the integration of GenAI.
We had a lot of that on display last week, all the way from Platts Connect and our ChatAI capability there, all the way over to iLevel with automated data ingestion, document intelligence on CapIQ Pro and others, many others, in fact. That is the first sort of core thing, making sure that we can actually, within the platforms that have the familiar workflows that our clients use and have used for a long time, integrate the technology there so they are getting that experience that they need there. The second thing is that we are and have been, I think, very front-footed and proactive in ensuring that we are getting our data into the LLM ecosystem and the hyperscale ecosystem.
This is a calculated bet in understanding that we can actually get greater license revenues as a result of that, but we can also potentially, over time, have an opportunity to pick up more customers through those channels as those channels themselves mature. You know, overall, we're very excited. To cap it all off, we announced Kensho Labs last week. We did that because our clients, you know, I mean, obviously we feature Kensho quite frequently when we're interacting with our clients. We've had really, really high interest in Kensho actually working directly with our clients' teams. You know, this will come from the highest levels, whether it is, you know, the business owners, the CIOs, the heads of AI, the heads of research, et cetera.
We already started with sort of a couple of, let's make an investment here and, you know, sort of see what Kensho can co-create with, you know, with this client. It sort of has started picking up momentum. We are formalizing it into Kensho Labs. You know, as we were saying to Andrew last night, it is both a means of us really just raising the stakes on engagement with our clients in ways that do give us additional opportunities to license new data sets and new capabilities. It also gives us a really wonderful, deeper understanding of what is possible in terms of where our largest clients are heading. That gives us some insight into things that we may also be able to do more broadly across our complete client base.
It's an area that we think there's a number of benefits from having that capability, and it's going over very well with our largest clients.
Okay. Great. You mentioned that LLMs like Claude for Financial Services is a place for additional revenue. You know, surely that could be a place where clients that do not use CapIQ discover CapIQ. But I could just see it like a client that might be on Claude for Financial Services, they might not need all of what CapIQ offers. How are you going to look at price for somebody who is not a customer of CapIQ, discovers CapIQ for a very, let's say, particular purpose, but is not interested in all of what CapIQ offers? Are they going to get a lower price if it is discovered on a third-party LLM?
Yeah. I would look at CapIQ as being side by side with the LLM. To the extent that we're not actually integrating CapIQ with the LLMs, we are integrating our data with the LLMs. From that perspective, it is not different to how we operate and have operated for many years now, which is that our clients oftentimes will license our data over multiple distribution channels. Oftentimes, even within the same client, if they're larger clients, we'll have one group in that client that licenses over one channel, one third-party channel, and perhaps another over another third-party channel. This is not at all uncommon for us. We've always taken the view that we would serve our clients where they needed us to be, including being very proactive early on with Databricks and Snowflake, for example.
This is another step in that direction in terms of being flexible with our clients. I think one of the areas that perhaps maybe the first thing to say is we're seeing a huge interest from our clients in figuring out how to use this data and think about how to use Claude more broadly. Because remember, it's not just for them. It's not just, here's another way to access data. It's actually, there's another tool. It's Claude. And how do we actually figure out how to integrate that and think about that? That's one thing. I think we are getting a tremendous amount of interest, more just in sort of understanding and testing. We've had a lot of clients basically tell us they'd like us to sort of quote unquote turn them on for their licensed data in Claude.
The other point that I would make is that we have had tremendous progress, which is perhaps a tiny bit counterintuitive in our energy client base through the integration of our content in Microsoft Copilot, Microsoft Copilot Studio over the past year. There has been a line out the door for energy clients to sign up to receive their data through Copilot, largely because Copilot is essentially, you know, a core part of their strategy. Many of these would not be clients that would invest in their own, you know, version of, like, let us say a Spark Assist, for example, internally. That has been very helpful. Perhaps a couple of points to make here.
The first is to the extent that we are offering an additional channel for our customers to use our data and also that our pricing around data is linked to the scope of usage, the amount of usage. In other words, like the use cases and the amount of usage, we would see opportunities to improve our economics around making that data available over more channels for our customers, particularly if we are integrating it with other data, which oftentimes our customers are now interested in doing over those channels. You know, the economics on this we view as an opportunity for us. We are also seeing that it's just increasing the appetite for using additional data as well.
I just asked you about that.
Yeah.
It's sort of like the same exactly you're hitting on. Have you been able to track your clients that embrace your AI tools the most? My question is, for that cohort that uses your AI tools the most, are they consuming more data from S&P?
Yeah. The appetite for data is significantly up across the board. It's not uncommon to be in a meeting now and have a client say, "What if we just had all of your data?" That's a very complex request from a client that, you know, we oftentimes sort of like break it down into, maybe let's try to figure out what problems you're solving for and sort of work with you to ensure we have the right use cases and that you're licensed for the appropriate things. We are absolutely seeing cases, for example, where we've closed some large deals through the CCO over the last month or so where just the fact of having our AI-ready data available to use in multiple additional new channels across the LLM ecosystems has been a core reason for closing those deals.
We've also had clients ask for additional licenses to data as they're testing out the LLMs. I would say an important example also, because remember, we do enterprise pricing. We're not an organization where our subscription products are based on, you know, per seat licensing. We look at the value we create and we look at that value across many different ways, including, you know, the overall usage and the value the customers are getting from the products. When we have licensed our energy data, for example, over Microsoft Copilot, what we found is that by using Copilot and just interrogating the data, customers have often had a kind of an aha moment where they didn't realize they were actually, you know, they had this additional data that was already in what they were licensing.
They are getting more value from that data by using it through Copilot as well. These are all ways in which we would expect to see opportunities either for improved economics or additional data licenses through these third-party channels.
Maybe we'll switch over to MI. Last week's analyst day, Sagata gave an MI outlook of 6-8% organic revenue growth for the space. I think financial desktops and data feeds broadly defined, it's got to grow more like mid to low single digits. And Sagata says, we assume we're taking share in 6-8%. So my question is, what gives MI the positioning to gain share, which in a place for financial desktops and data feeds, it feels like a multi-provider market?
Yeah. So I would say that, you know, the two components to Market Intelligence are, if you remember how we sort of broke it down, we have the data and the insights components of it, and we also have the key workflows components of it. You know, obviously our 6-8 spans both of those. When you think about them, you can certainly look for our subscription products. You know, you'll have seen, for example, in Q3 that our subscription, our ACV revenue growth in Q3 was 6.5-7%. That is already in line with the medium-term guide that we've provided here. We have done a tremendous amount of work, as you know, around revenue transformation.
I just talked about all of the innovation that we're doing on the products by integrating generative AI, and we are bringing in additional data through WIS Intelligence, for example, and the partnership that we announced with Cambridge Mercer. On the other side of the business, you have the fact that we have incredibly core workflow tools in many of these areas. Some of them we named out last week, whether it is Wall Street Office, iLevel, Notice Manager, and many others that we operate in Market Intelligence. Those businesses are growing very fast, which you can obviously see from our results as well. You know, those two coming together give us good confidence in the outlook for Market Intelligence from the perspective of the medium-term plan.
We're very happy with the progress that we've been making on the data and insights piece with respect to vendor consolidation as well. That's an area where we've been, I think, a net beneficiary over the last couple of years also, I think.
You're still being disciplined on price, right?
100%. Yeah, 100%. It's really about, I think the conversation with our clients is very much about, look, we can help you to reduce your overall vendor cost by consolidating to us. You know, rising tide lifts all boats, essentially. That has been helpful to us as well.
Makes sense to me. When you're looking at the medium-term algo EPS growth, which is low double digits, which is a little bit lower EPS growth than, let's say, the last three years, I think the simple answer to it is you're just starting at a higher margin now. I think that's really kind of the answer. My question is, if you look at S&P and your margin is now higher than three years ago, you're probably closer to your incremental margin. That's why we're really talking about low double digits instead of teens EPS growth over the next few years in the algo. My question to you is, what are you doing to raise the incremental margin of S&P?
Yeah. Very well said in terms of how we're approaching this. I think just generally, we obviously have the tremendous benefit of having transaction businesses that both can grow at very steady rates, but also have opportunity for outperformance at various different times for, you know, maybe it's a rapid expansion in a particular asset class or, you know, rapid outperformance in equity markets. We always have those opportunities to outperform. I think that's a relevant statement when it comes to the medium-term plan that we've laid out here for both margins and EPS.
I think what's, you know, the core of what we're thinking through here in terms of margin expansion, EPS growth is being able to take the scale that we're creating and the operating leverage that we're creating, whether it's in our data, our technology, and some of the core areas such as research, for example, across the organization, and actually be able to turn the productivity levers there, ultimately with the goal of growing revenues without having to grow headcount. We've certainly achieved that in the Enterprise Data Office. We look to do that across other functions as well as we scale. You know, that's where we, you know, that's where we have the tipping point really for the organization. We have a lot of initiatives in place, some of which Eric and I touched on last week around this.
A lot of that is also centered on accelerating the integration of generative AI.
Maybe you can say, I just answered this question, but let's jump into efficiencies at ratings. When you look at current margins, I mean, 2025 margins, they've surpassed back the 2021 margins. And back in 2021, the company said, oh, margins are really high because, you know, of course, issuance is high. Now it seems like it's really kind of more normal costs and rising from here. What's happening in ratings margins where, you know, they're rising in a year where, you know, issuance is good, but it's not driving the whole margin story this year.
Yeah. We've been, I would say since 2022, very deliberate in streamlining our overall processes within ratings. Some of that was to do with some of the changes, the organizational model changes that we announced around the analytical leadership structures that just brought teams together and made it easier to work across the organization. We were speeding time to market and, you know, reducing the burden essentially on some of the core teams for delivering the ratings. There was also a very definitive effort to integrate technology. It started with the more routine RPA or robotic process automation and machine learning. Ultimately, obviously now we are into integrating generative AI at scale across our analytical workflows. That has been extremely helpful in, and again, in making sure that we can do more with the same.
In some cases, you know, for some operational roles, for example, we were even able to do more with less. I think we were also quite thoughtful about how to think about analyst expertise and capacity. Of course, we have invested in the fast-growing areas, whether it's CLOs, asset-backed finance, and infrastructure. We've also done a lot of cross-training and recertification of the analysts in core areas so that they're actually fungible and can move between sub-asset classes, for example, within structured finance as needed so that we can continue to have the capacity that we need there without needing to, you know, to add additional headcount. We can sort of load balance, if you like.
Let's touch on private credit, which is a huge asset class. I surely know that S&P is well positioned to capitalize within both Ratings and MI in serving those end markets for private credit. At this point, and I know it's been a fast-growth area for you, given the size of the private credit market as it is today, wouldn't you think that you would have even more Ratings and MI revenues from this, you know, kind of mega trend towards private credit?
We're actually very, very pleased with the growth that we've had across our private markets franchise. Of course, we've been sharing that information with you over the last several years. The growth rates in ratings have been very, very strong for private credit. A lot of that is down to the investments that we made, not just in the analytical capacity, but also in getting out and engaging the GPs and investors around our methodologies and around the ways in which we would interact. I think we've seen some very, very strong growth there. It's across all asset classes. As you know, this isn't just about one high-yield, you know, borrower going from, you know, either the high-yield bond market or the leveraged loan market into private. It's really spanning now across asset-backed, structured finance, middle market CLOs, infrastructure.
We are very, very well positioned and very pleased with the relationships that we've developed on the commercial front there as well. I think across the rest of the organization, we see really significant opportunities to have very accelerated growth also. Some of that just comes from the connecting together of current and new capabilities that we would expect in Market Intelligence, for example, with the announcement that we made with Cambridge Associates and with the acquisition of WIS Intelligence. Each of those brings very unique elements that help us to what we call, what we would call, say, close the transparency gap or solve the transparency gap in private markets. You know, we are looking to provide benchmarks to provide reporting in a consistent way, consistent taxonomies.
Our Cambridge Mercer partnership gets us to a common taxonomy that Cambridge Associates and Mercer developed with us for reporting. They also bring contributory data as well as with other contributors. WIS Intelligence brings some very valuable contributory data and quite unique data on investor preferences as well. I think the combination of these things really positions us ideally. Our view ultimately is that this market is very large. It is deep in terms of the opportunity for us to serve. There is a massive demand there for the types of services that we provide around transparency, whether it is ratings, valuations, or benchmarks.
As much as you've been successful, I assume you're saying there's no question this is a large opportunity ahead.
Yes. Yeah. It's one of the largest opportunities for S&P Global over the medium term.
You mean that both in MI and in ratings? My question is, is that embedded into the algos that you gave at Allen State last week?
Yes. Yes. Also in index. There are, albeit starting from a, you know, kind of a smaller base, there's a tremendous level of interest in index as well. The index team has been innovating and launching indices this year. We have some really exciting things lined up there also.
Questions from the audience? Come on. My team is right in front of you. Someone has a question. Go ahead. Speak to your first-year CEO.
Yeah. Can you speak to your first-year CEO? I am curious where you feel you are in terms of having the talent around you to execute? Because the business is changing, the go-to-markets are changing. I am just curious, how is the organization changing with that?
Yeah. Thank you. I'm, I would say pleased, of course. You know, we're always our own harshest critics. So there's always things I would love to do more of, faster, you know, better. I think the team has performed exceptionally well. I'm particularly proud of, you know, the division performance. Market Intelligence, I know, has been an area that, you know, that has been looked at and examined closely for a few years. I think we're essentially building to and legging into, you know, to the way that we expect to operate, you know, going forward. Very pleased with that. I guess the other point I would make is we've, we set out with the Enterprise Data Office and the Chief Client Office a very, very new way of working. Learned a lot.
Actually got a lot of wins out of both that I referenced earlier. We see opportunities there within technology, for example, to do some similar things. Our technology leaders are very excited about that. As it relates to, you know, sort of like the overarching leadership team, I'm extremely pleased with the team that we have. We obviously have one open role with our technology lead, but this is the team that will take us forward into the foreseeable future.
Okay. I think we should conclude. Martina, thank you. Thank you. Appreciate it.