Good morning and welcome, everyone. Thanks for joining us at our 42nd Strategic Decisions Conference. My name is Kelsey Zhu. I'm the Information Services Analyst at Autonomous. With me on stage today, I'm very pleased to welcome Chris Cartwright, the President and CEO of TransUnion.
Thank you. Good to be here.
Thanks so much for joining us again.
Always a pleasure.
Chris, I know we have a lot to talk about today. There's macro, there's consumer, there's AI, there's the whole credit score transition topics I want to pick your brain on. I think a good place to start is, you've obviously had nine consecutive quarters of delivering organic revenue growth of high single digit and above. In the last quarter, your growth actually accelerated to 11%. This is definitely outperforming the type of credit volume trends we've seen in the market. Could you maybe just tell us a little bit more about what's structurally different in TransUnion's portfolio today that's driving all the strong performance?
Just a bit about TransUnion, because I know there are some generalists in the audience, and I want to level set before I answer your very good question. We're one of the three large global credit reporting and consumer information companies, if you will. We operate in 30 countries around the world. Roughly 80% of the revenue comes out of the U.S., where our business is a mix, like half credit, and then half the combination of marketing, fraud mitigation, public records, and communications solutions. We have been undergoing a transformation the past four years along two key dimensions. The first, through a series of acquisitions that we made in late 2021, early 2022, we significantly expanded our product capabilities. Obviously, we're steeped in credit and credit analytics tradition. Now we're also very good at digital marketing enablement and online and phone-based fraud mitigation.
All these services are unified now on a common technology platform called OneTru, and it's really that technology innovation and standardization that's the second leg of our transformation. It's going to allow us to change our operating model from one where we have independent technology stacks across these 30 countries that are essentially doing very similar things for the same industries, for the same customers with common information, to a global configurable platform, which does two things for us. One, any country, or as we migrate countries to that platform, suddenly all the services that it enables become available, many for the first time in those countries. We're launching new growth paths, if you will.
Secondly, we're getting cost scale because we won't be doing the same things redundantly around the world, which was really just a function of being a 50-year-old business that expanded country by country, one bank consortium at a time. Now, you mentioned that we've been growing well, and our growth is accelerating, and it's above market volumes. That's all true. I'd say it's a function really of a couple of things. One, really, since we became a public company in the summer of 2015, with only a couple of exceptions, this business has been able to grow high single to low double digit organically.
The only exceptions to that would be 2020, the COVID year, followed by 13% growth and a full recovery, which took us to kind of peak lending volumes in the super low interest rate environment, the ZIRP environment as you folks call it, before inflation spiked dramatically in Q2 of 2022. We had an aggressive Fed response, 500 bps of rate escalation, which put consumer lending volumes in a recession for the next six quarters, if you will. During that period, we grew 3%. In 2024 and 2025, we accelerated again to 7% and 8% organic growth ex FICO. This year we're trending hopefully as good or better. We'll get to some of that update later.
The point of taking you through that brief history is to say that this is a growthful market, and the data that we're providing and the analytics we're providing is dimensional and growthful, and can sustain that level of growth independent of the second reason that we're growing to the top of the market is we now have a lot more things to sell the market segments that we serve, whether that's our financial services market in the U.S. or a range of what we call emerging markets like telecom or technology, retail, e-commerce, the media segment, et cetera, et cetera, et cetera.
There are over 20 different market segments that we organize ourselves around in the U.S., and now we can sell them credit, marketing, fraud, identity verification and resolution, direct to consumer credit enablement, investigative solutions, whole range of different products. It's a function of growthful markets, product diversification, solid tech execution.
That's a really helpful summary. Thanks so much for that, Chris. I guess a good place for us to start diving into some of the macro consumer debates is really just I feel like we've been having conversation around a K-shaped economy for a while now. What's the latest in terms of the health of U.S. consumers and specifically for subprime categories? Could you just give us a quick update what you're seeing in the market?
For sure. Happy to. The K-shaped economy is real. The middle is transitioning. However, it's not necessarily a negative story or nearly as negative as is often portrayed. We recently just produced some analysis that shows that the primary driver of this K shape is that folks in the middle of the credit spectrum are migrating upward and becoming near prime, and super prime. It's less that they're deteriorating and ending up in the subprime. The percentage growth at the bottom of the K, people transitioning from the middle to the subprime, is materially less than the percentage growth of consumers whose credit is improving, and they're getting structurally more stable. Now look, it's a different question about whether that kind of divergence between lower- income, higher- risk borrowers in the latter is good or bad.
I think oftentimes there's the sentiment that it's all negative and consumers are moving from the middle to subprime, which in fact, they're not. Over the past quarters, we've looked hard at the performance of recent loan originations within the subprime space, and their vintage curves, their performance curves are very consistent with the prior year. There's been no concerning increase in loan delinquencies. There's been some modest tick up, but you would expect that given that lenders expanded their loan boxes and made more loans into the subprime space last year than they did previously, as their confidence in that segment increased and continues to this day. Lenders are managing their risk in large part. They're taking on more customers in the subprime space. They're making a higher volume of loans.
They're also making a lot of small- dollar bets to determine the consumer's ability to service the debt over time and to mitigate their own risk. We don't see any red flags. We don't see any concerns with the performance of loan portfolios amongst subprime borrowers at this point.
Consumers are relatively healthy still. We're still in this benign credit market.
I definitely understand the macroeconomic concerns. There's an enormous number of uncertainties, which we'll discuss, but consumers are still highly employed, and there are still some real wage gains above inflation. Of course, inflation may spike. We don't know. That's a big part of the uncertainty and anxiety. Through April and through May, when I look at the performance of our business, our portfolio throughout the U.S., it's completely consistent with how we've been guiding investors year to date, from the initial guide for the full year to the first quarter earnings results. We're very much on track to deliver what we guided to.
Got it. You mentioned that some of the lenders are expanding the credit box a little bit. I was wondering if you can talk a little bit more about consumer credit, origination volume trends that you're seeing across different products. Also, just the environment we're in right now. I don't know if it's correct to call it higher for longer or hold for longer. Are lenders still expanding that credit box, or how are you thinking about the impact of higher for longer on your portfolio?
Sure. Well, look, as I mentioned, and as you know, we outperformed our guidance, the high end of our guidance, substantially in the first quarter. Because we saw these uncertainties building and we wanted to be very confident with investors that we could deliver the full year guide, we chose to bank the overperformance and increase our contingencies, if you will. Those contingencies were already adequate because we started the year by issuing what we call a prudently conservative guide.
We were also very clear in the first quarter earnings that if market volumes across these lending categories held, we would expect to overperform the high end of the guide to some degree. We don't see any changes in the lending volume across card, auto, consumer lending, even the fintech component and mortgage that would prevent us from doing so really through the middle of May.
We're feeling pretty good about that. Now, of course, there have been any number of macro uncertainties in the recent year or 18 months. There have been concerns about the inflationary impact of tariffs, obviously. That didn't nearly manifest itself to the degree that some had thought. Concern about increased levels of deficit spending. That proved to be pretty manageable, and inflation was down to the high 2s, maybe 3 level. Of course, the conflict in Iran and the lack of oil flowing through the strait and all of that is a whole another different contributor, and we just have to see how that is going to manifest itself. I guess the point is, sitting here in late May, it has not manifested itself negatively in our business volumes to any degree.
Got it. Stable volume trends is what we're seeing.
Yep.
I guess one of the main investor debates or conversations we've been having this year on information services companies is really the AI impact on the whole sector. I do want to talk about AI disruption risk and competitive landscape, and things like that. I think oftentimes people don't really ask about or really debate about the AI upside to some of these information services companies. I thought a good place to start there is just in your last investor presentation deck, you highlighted AI-enabled customers are actually consuming more data, and that's a positive impact on TransUnion. I was wondering if you can talk us through some of that, what you're seeing, and the potential revenue uplifts that you could see from this AI implementation.
Yeah, for sure. Look, as a first principle, I think we can all agree that AI can do very powerful things with access to quality, curated, relevant information at scale. That's no different in my part of the information services industry, credit reporting and marketing and fraud services and the like, than anywhere else, legal research or market information or medical publishing. The difference is that the information that we provide as our business foundation is very difficult to access. You need to have tens of thousands relationships in the U.S. market. Those are individually contracted.
Those are managed on a monthly basis ongoing. You have to have a legal permissible use to gather the information, and then you have to be very careful about who you allow to access it. You have got to monitor the manner in which they're using it.
This all takes place under a very tight, clear, and punitive legal and regulatory framework. I think that creates a considerable barrier to entry because you simply can't get access to the credit information for the AI to do what the AI does. The other problem is that generative AI isn't really applicable to core credit decisioning because it's legally required that you be able to expose the full audit trail of how a loan decision was made and inform the consumer of that, and that's not possible with generative AI. However, you can still leverage AI and machine learning in order to accelerate the development of predictive models. That is really the basis, the purpose of all of this credit information. As I look at our business, it's a very large and defensible moat around credit information.
There is a similar moat around the marketing and the fraud data that we provide because much of that signal comes from broad global industry consortiums that we've built with hundreds of companies, and we have years and years of data there. That's a difficult barrier to access. AI principally gets its information in the foundational models by crawling the Internet and crawling publicly available sources, and doing broad content licensing deals and the like. None of the information that we provide is really accessible in that way.
I feel like first, the data layer, the intelligence foundation for our business, if you will, exists behind very material barriers to entry. The purpose of the data is to use it to build models that can better predict the future, better predict outcomes. Outcomes for loans or marketing campaigns, or certain fraud prevention techniques.
We can employ that type of technology to more rapidly refresh and deploy models in much less time, at much lower cost, which is going to allow us to do more of it and better service the full spectrum of our clients. Now, when you think about a bureau like TransUnion in the U.S., we have thousands of clients. Some of them are very large and sophisticated, and they really just need our data and understanding about what data is available for their own modeling and internal efforts. We call that the do-it-yourself segment of the market. There's an equally large segment that need our assistance. That's the do it with me segment of the market. There's a third equally large segment that really wants our help, wants us to do it for them.
Those three growth segments, there's an opportunity for us to really expand the range of services that we can provide them based on the acquisition and the integration of all of these products on a common platform, but also the depth to which we provide it. Right now, a typical lender may only refresh their loan origination model once a year. It's difficult, and it is somewhat costly to do it. With AI and with the tech innovation of bringing all of our data together on a common platform, we can refresh those origination models or the cross-sell models or delinquency and collection treatment models, whatever aspect of the lending life cycle you want to focus on. We can do it more frequently and more effectively in a forward- deployed partnership with our clients.
It's going to enable us to really change the manner in which we serve our clients from more reactive and at their request to proactive and enabling. That's really something that wasn't possible before AI and the application of it across our datasets.
Got it. There's a lot to dive into there. At TransUnion, when you think about your AI implementation initiatives and processes, I guess how much of that is focused on internal cost save opportunities? How much of that is focused on externally driving, whether it's new products, new client segments that you weren't able to previously in the pre-AI world? Maybe talk us through some of the initiatives that you're really excited about across credit, marketing, fraud, consumer.
Well, first, just as we were discussing, I guess I'm probably most excited about the ability to use our data and AI in combination to service our clients' credit fraud, marketing, identity resolution, data hygiene needs more proactively, more holistically than we could previously, and really kind of transform the frequency and depth with which we do that. Right? I think that's an enormous revenue growth opportunity for us. You were at our recent Investor Day. A number of investors commented to me that they've been to the website, and they've seen demos of the agents that we've created to orchestrate and accelerate analytics and model development. We can now build all those predictive models in a fraction of the time that it took previously.
Our large analytics consulting organization, which is a key enabler of our data usage, is orders of magnitude more productive. That's where they're going to become more forward-facing, more kind of forward engineering and consultative than before. I think that it will enable a lot of growth because, look, the truth is our business and our industry has evolved a lot in recent years. We're not just about credit. We're a broad global diversified consumer data and insights companies. Most of our clients don't appreciate all that we can help them with. This is a great way to both evangelize the things that we can do, but also deliver real value analytically. That's the first thing AI is accelerating that I'm super excited about. Now, obviously, software development and data science is one of the most critical things that we do.
It is our internal manufacturing process, if you will. I'd say 60% of my global headcount is in those two areas. Those associates are more productive because of AI support in the whole spectrum of software development activities. We estimate between 20%-50% productivity improvements across our technology organization, depending on what they're doing and how well AI can enable it. That's definitely true on the analytics modeling side, where things like our analytics orchestrator are driving orders of magnitude productivity increases. Now, we're in the early days of applying artificial intelligence to some of our consumer and customer support operations. I do think eventually you will see nice savings flow from that. When I think about savings and the potential for them across the portfolio in the medium term, AI will be an important part of that.
Remember, because we've been investing on the technological side to build a global product platform, OneTru, and most of our business will run on OneTru in the medium term, that's going to allow me to dramatically reshape my product, my technology organization. Where today I'm running on separate individual technology stacks in each of the 30 countries that I operate in around the world, that's going to be replaced by a global configurable tech stack enabled by all of our proprietary data, and I can deploy that in markets around the world. Our business model is going to evolve from the sum of vertically integrated multi-domestic pieces to one that is globally architected and run on a global product platform, but also an internal operational platform that's shared across our business in different markets.
Just deploying all of that, which we've created over the past four years, and we're in full rollout mode, that means material structural cost reductions that we'll use to accelerate revenue through sales deployment and new innovation, but also flow through to bottom-line margin enhancements. In addition to that, we're going to apply AI.
Have you thought about the sizing of that medium-term cost opportunity that you highlighted? One of the key stats that you mentioned is 20%-50% uplift in productivity for developers.
Sure.
It's also a question of how much of that do you reinvest, how much of that do you give back to margins? Maybe just tell us a little bit more about how you think about that.
Yeah. The blended estimate on improved productivity for my developers is, say, 30%. If you talk to any CEO in information services, they will tell you that either their innovation pipeline or their internal efficiency pipeline that requires software development to fulfill it, ideally needs more than a 30% productivity boost. It's great that they're more productive. I can absorb all of that productivity and drive faster innovation and faster innovation deployment and a better operational tech stack and use that to become more efficient, rather than downsizing any employees, right? I'm going to use it as a productivity enabler and then implement those benefits. Look, we had an Investor Day recently. The medium-term guide calls for improving margins 50 basis points a year over the next three to four years.
That's in addition to a 240 bps improvement in margin that's transpired over the past four years, that was largely masked by the price increases that FICO was putting through that were included in our revenue. Again, that's revenue upon which we have no margin. It was masking the operational leverage and the revenue flow-through that we were seeing in the business. Now that we can break that out, investors can see that indeed our margins have improved a lot over the past four years.
To get to really what you're asking, can I outperform the 50 basis points? What I would say is this, putting uncertainties aside, if I could count on 7%-9% compounding revenue growth, and a lot of that because it's coming through with a good mix of credit, I have a lot of flow-through to profit. That, coupled with the structural cost reductions, makes me highly confident in the 50 bps as a floor.
Got it. Super helpful. We'll come back to FICO in just a second. Maybe one last question on AI, and I know, Chris, you touched on this at the beginning of this AI discussion about your clear competitive advantages in the credit space.
Sure.
Maybe tell us a little bit more about how you think about AI-native competitors in fraud, marketing, consumer, and how are you thinking about TransUnion's competitive advantages there?
Yeah, absolutely. I grew up in information services. I probably have worked for 25+ years in different information services companies or conglomerates that had legal, tax, financial, medical, healthcare, data, et cetera. I keep track on what's going on in a lot of these spaces. Again, as a rule, where the information is broadly accessible, then new AI competitors can come and challenge the competitive dynamics in the industry. You're seeing different variants of that as you look broadly across the space. That's not to say that the incumbents that have a massive advantage in terms of the data they've got and their domain expertise won't respond and compete very effectively over time. You don't see that dynamic in credit because the moat around the credit data and its mission criticality and the high regulatory content is so substantial.
In the marketing world or in the fraud world, there are various AI-enabled competitors that have been there for quite some time. I'd say the advantage that we have as an incumbent is all of the proprietary data that we get from the industry consortiums that we've created in fraud and marketing. In fraud, we've got this decade and a half consortium of companies that use our software to detect potential fraudulent behavior on their e-commerce servers, and then they report that to us, and we have a large device reputational network that is proprietary and is a considerable moat to providing online fraud detection. In the phone world, that data comes from the carriers. We're one of two scaled players in providing phone authentication.
Any other flavor of fraud mitigation, be it biometric or knowledge-based authenticators, et cetera, we either originate market-leading solutions in those areas, or we have a full range of partnerships, and we take all of that signal of potential fraud, and it goes into the OneTru Data Foundation, and our clients can build rapidly, very sophisticated AI-enabled models to score the likelihood of fraud in a transaction. In short, it's the combination of our architecture and our analytics sophistication fueled by a lot of proprietary data, even in areas like fraud or marketing, that gives us an advantage against new players that come up that are really just leveraging somebody else's AI capability.
Got it. So scaled data.
Scaled proprietary data, modern tech stack, AI sophistication, deep client relationships, forward engineering organization in a massive domain wrapper.
Got it.
Look, when people think about AI and the potential for disruption, they often come at it first from, "Is this theoretically possible?" The answer is, sure, it's theoretically possible. The second question should be, is this economically viable, given the advantage of entrenched players and their technological sophistication and the clear motivation we all have to be excellent at this? I think there's a lot of fallout when you consider economic viability against large, sophisticated, entrenched competitors.
It's very well said. Thanks so much, Chris. I guess switching gears to talk about the credit score transition that's taking place right now, maybe a good place to start is just the VantageScore pilot program that FHFA Director Bill Pulte has announced a few weeks ago. We heard about these 21 approved lenders. We heard various different things about the Loan-Level Pricing Adjustment for VantageScore. I guess from your seat, what are you hearing from lenders, MBS investors, regulators? How is the pilot program going?
Sure. Yeah. Well, no shortage of things to talk about in the mortgage realm, particularly around the implementation in earnest of score competition, which I think is a really great and healthy thing for the U.S. mortgage market. Simply because the VantageScore is a next-generation score based on trended and alternative data and developed with modern scoring techniques. It's a very good score. It performs very well against the incumbent. The challenge, of course, is that current processes are designed in consideration of only one score for over 30 years.
The way I would characterize it is we're in the early days of working through all of the adjustments, all of the learnings, all of the human behavioral change that has to happen to enable share shift and competition at scale. The good news is that the FHFA, this is for real, this is happening.
There is massive interest amongst lenders. In the 21 are some of the largest mortgage originators in the U.S. People like UWM, people like Rocket Mortgage, a variety of the largest banks, et cetera. There's just a tremendous proven tangible appetite for change given the massive price increases that have been put through by the incumbent, not only over the last four years, but really since 2006 forward when Vantage was formed to create a counterweight to their monopoly. Right?
There's just no doubt that there's market demand for this. There's no doubt that the score is better, and of course, it's going to take some time to work it through. That's all happening because there's a tremendous economic incentive on behalf of the lenders for it to happen. Score improvements contributes to overall safety and soundness as well, which the FHFA wants to happen. Most importantly, the FHFA is also very focused on reducing loan transaction costs. A score that costs $0.99 versus one that costs $10 or $5, plus a contingency on a successful borrower, these are material dollars that can be taken out of the system that aren't contributing a lot to the economics.
I guess what are the remaining hurdles for a full-scale implementation for VantageScore? It's great that there are 21 approved lenders, but there are hundreds of other smaller lenders in the market.
For sure.
I guess what's your expectation around both VantageScore full-scale implementation timeline and maybe FICO 10T implementation timeline as well?
Well, the FHFA has said that they hope to certify FICO 10T over the summer at some point, and we'll see if they can deliver on that timeframe. I expect that FICO 10T will be a materially more performance score than Classic FICO because it takes advantage of the alternative data and the trended data that the bureaus provide. Ultimately, it's the data that determines the efficacy of the score. It's not the magic of the score.
That's what we understand on that development. I think we're in a prolonged period of transition. This is my estimation based on what I gather from a variety of sources. This isn't some definitive proclamation by the FHFA, but the FHFA is working at this in earnest. The GSEs are ready and will probably take the majority of this year to continue to educate and refine the process. There are multiple players in this mortgage origination through securitization process. They have to be brought along. There are some systems adjustments that have to take place. This is a large change management process, and it's going to take some time.
It is.
I think that's why it was prudent that each of the bureaus came out and assumed no adoption of Vantage over the course of this year and said, "Let's make 2026 a year about learning and change management." I'm sure some of that will continue into 2027. I do think there will be some share adoption of Vantage in 2026. You're seeing that. This is not a flip the switch. This is not a single event becomes the aha moment that proves that nothing will ever change. I think those are all kind of wishful and unrealistic assumptions. The fact is, it's going to change. There's going to be score competition. There's a tremendous economic incentive to adopt Vantage, and it's just going to take some time to work it through.
Got it. I guess in the non-mortgage verticals across credit card, auto, personal loans, what type of market share has VantageScore been able to gain over the years? What's your medium-term expectation for VantageScore market share gains in mortgage?
Yeah, look, that's a good pivot because oftentimes the FICO versus VantageScore discussion is framed in a way where it's as if Vantage is this new thing that has just appeared on the scene. Vantage has been around since 2006. Vantage is significantly used in consumer lending. There's a major card issuer, Synchrony, that for not quite a decade now has been fully on Vantage originating, securitizing. You see it in auto as well. We've had a nice business amongst community financial institutions that originate and hold a lot of mortgage moving to Vantage because of price escalations from the incumbent. Right? It's known. It's used across 4,000 banks. It's the most frequently pulled score or offered score in consumer education.
What I think will happen is: I think the share that exists outside of mortgage exists for a large reason because of the halo effect of the monopoly that exists in mortgage. As that monopoly becomes competition and share begins to balance between the two players, I think it will have a knock-on effect on share dynamics in these other lending categories.
I guess one of the main initiatives or projects that Director Pulte is really excited about is just reducing closing costs. Under that or in this general operating environment, how should we think about pricing for VantageScore and credit data in the next few years?
Well, just focused on the score for this time. Look, we operate in countries around the world. We produce scores. There's some other third-party scores that we would distribute in one market or another. Around the world, there are no other scores that sell for the price point of a FICO score in U.S. mortgage markets. Right? I think that's just a function of a lack of competition. I'm very comfortable with the score where we're at. I don't think this is a time to worry about pricing or revenue realization. It's let's be supportive of industry change. Let's lay the foundation for real competition going forward and enable share shift. Over time, look, any revenue we get from a score is additive to our P&L currently. Within our business, we have fully decoupled our data revenues versus third-party scores. Right?
FICO's at liberty to manage their business as they see fit. We're more than happy to calculate the score on our data behind our firewalls if that's what the client wants. We're more than happy to support FICO's direct licensing program. If I could wave a wand and 100% of FICO score calculations shifted to resellers or clients, I'm fine with that. My economics, my revenue, my profit is unchanged. It's just my margin goes up a lot because I no longer have to book zero margin FICO revenue in my P&L. I think it's important to understand that decoupling that's been achieved over the past year or two. Again, we're focused on share shift and share adoption. Pricing will remain what it is today for the long foreseeable future.
We will continue to take the very reasonable price increases that we have put through both TransUnion and our bureau competitors that really reflect significant investment in innovation. Remember, just recently we had Director Pulte and Director Turner of HUD celebrating score competition, celebrating the use of alternative and trended data. You have to remember that that moment was made possible by 20 years of patient investing by the bureaus in creating a credit score alternative and creating trended and alternative data. The fact that we have put through some price increases and will continue to do so on our product reflects a lot of underlying innovation and investment that's taken place over two decades to get to this point.
Got it. Maybe just one last question on mortgage.
Sure.
I do want to touch on the whole tri-merge, bi-merge, single file underwriting debates. We've seen, I think, the FHA has already said that tri-merge requirement remains for the time being when we're going through the credit score transitions. I guess worst-case scenario, if we do move from tri-merge to either bi-merge or single file underwriting, could you just talk us through TransUnion's competitive advantage in the credit data space?
Yeah. Look, there's a number of ways to come at this, but I think we have to start with likelihood and severity. Right? Because investors get concerned when they hear regulators make comments that, while not about the tri-merge necessarily, raise some question about further potential change. It's my belief, it's our belief based on any number of discussions we've had in and around Washington and in particular within the FHFA, within the GSEs, within the HUD, within the Treasury, within the administration, within Capitol Hill, across Capitol Hill. There is firm and broad support for the tri-merge. At this point, really the only people who are advocating a change is the Mortgage Bankers Association. I appreciate their circumstances. The past five years, it's been very difficult to be a mortgage broker or a mortgage originator.
Volumes are dramatically below the curve. I understand why they want to try to find cost reductions wherever they can. It's imprudent to go to a single pull or even a bi-merge because you lose a lot of important data, a lot of predictive signal that helps you. One, determine fraud and two, most accurately price the mortgage for the consumer that you're trying to win, and then ultimately to the GSEs to whom you want to sell the loan. There's a strong economic reason for lenders to pull multiple scores in the tri-merge. There's a strong overall safety and soundness argument in support of it.
Let's say, even if it does happen, right? Well, look, we love our mortgage business. We love the tri-merge component. We value all of the components of our revenue stream. It wouldn't be that material to our overall P&L, right?
We will be over $5 billion. We will be approaching $1.9 billion in EBITDA. When you look at the component of our business that is from the tri-merge at underwriting, it's a relatively modest portion, particularly because we're not talking about it all going away. We're talking about the potential reduction from three to two and perhaps some price competition. I would just remind everybody, we just went through this over the past two years, right? The FHFA said you no longer have to pull at qualification, three credit reports in order to get a signal from the GSEs about whether they would buy this conforming mortgage; you could just pull one. The market has, over two years, reached an equilibrium of pulling more than two. I would say that our mortgage revenues have grown very nicely over this period.
Our share is stable to growing, and I don't think investors have really noticed the change. If the same thing were to happen, and again, it's a super low probability at mortgage underwriting, when I don't think it's nothing that we wouldn't grow through in a single year.
That's very helpful. Thank you. Switching gears to international markets. The two markets I definitely want to touch on is Mexico and India. In Mexico, you just completed the acquisition of TransUnion de México.
Yep.
Maybe tell us more about what attracted you to that asset in the first place, what kind of revenue or cost synergy are you expecting, and how should we think about the medium-term growth algo and margin profile of this business?
Well, from the get-go, Mexico will be accretive to our overall growth rate and our overall profit margins, even as we're integrating the business this year and not separating out integration costs, right? It's a very profitable and attractive market and asset, and it's one that we've been in for over 25 years. We owned 26% of the bureau over this period. We wanted to buy fully into it, but the Mexican banks wanted to retain the asset at that point, and we were their technology partner. Finally, circumstances presented themselves where we could acquire the bureau in full, and we have. The way it stands today is it's a very basic bureau. It's got good data coverage across the large banks and the smaller banks, and the fintechs as well.
We will materially broaden the coverage, and we will re-platform it from the legacy tech stack, which is considerably behind OneTru, which will bring the range of our credit analytics sophistication into the Mexican market, coupled with all of our marketing, fraud, consumer enablement, and identity resolution capabilities, all of which are native and integrated to the OneTru platform. In short, we're going to enable a lot better credit functionality plus a whole range of additive complementary functionality in the Mexican market, and all taken to market with the thought leadership and the partner problem-solving orientation that we have in every market in which we compete. Mexico is a big market. It's fast-growing. Traditional financial services are under-penetrated. What we do, both on the data and analytics side, is a real enablement for economic growth and financial inclusion.
I think Mexico will become one of the real bright spots in the portfolio over time.
Speaking of bright spots. I guess India has been a key growth driver in TransUnion's portfolio for so many years.
Yep.
I guess sometime in 2023, I think the end of 2023, we had the RBI's cooling action, and now we're just getting out of this bottom of the cycle. Maybe tell us a little bit more about how you think about the recovery trajectory for TransUnion's India business in 2026, 2027, and what level of new normal in terms of medium-term growth rate can we expect for that business?
Well, sure. Well, look, first things first. The good news is we are recovering in India in the second quarter, as we expected and as we guided to, which puts us on track for the full-year recovery that we're forecasting in 2026. You characterized the RBI actions as cooling actions in 2023. I think that's exactly right. We had been growing at about 30% a year for three years. It was fueled by a lot of economic growth, a lot of lending activity, particularly online, unsecured, new-to-credit lending activity. The RBI became concerned about some of those practices and perhaps that it was a bit overdone. The good news is, despite the cooling measures that they put in place, which actually worked really well, much to our volume detriment, the delinquencies of the loans that were originated in that period have been very stable.
It didn't manifest itself as a lending problem, a loan problem, but it definitely cooled the market. What you saw replacing it, and what has helped that market return to growth, are gold loans, which unfortunately don't require a lot of credit pulls, although we're starting to introduce some analytics around that. Aside from that, personal lending, card lending, et cetera, has bottomed and is starting to grow, and that's why we're starting to see the improvements in our results. I feel like you're going to see some good organic growth over the remaining quarters of this year that get us back on track. Obviously, India is a fantastic market to have this fantastic and large share business in, and I would expect medium term guidance, it's mid-teens plus.
I don't know that we have provided medium term guidance necessarily on India, but I'm loathe to get too far ahead of the next quarter, because I just want to start to see more of this recovery. Look, things are still strong in India: a massive, productive population, under-penetrated in traditional financial products, a sophisticated regulator in government that wants to build a world-class financial services industry, and a heck of a lot of GDP growth. I'm thrilled with that position.
Got it. Maybe just one last question before we wrap. How are you thinking about capital allocation?
About?
Capital allocation between buybacks, debt payment, M&A?
Sure. Well, look, our debt is in very good shape. Even post-acquisition of Mexico, we're at about 2.8x . We will easily go down beneath our target of 2.5x leverage by the end of this year. We really want to focus a lot of energies on share buybacks. Unfortunately, the growth of the hyperscalers and the frenzy around all components of the AI stack is sucking a lot of oxygen out of information services. Some fear of AI displacement is also impacting information services, and we want to buy back shares to the greatest extent possible. I expect that we will buy back roughly the same amount of shares this year as we did in 2025, where we started to accelerate it, but that's only because we had to pay for the Mexico acquisition this year.
Next year, we've got all of our free cash flow, which is now solidly above 90%, and if we're still in this position where there's a big share value dislocation, we'll be aggressive in buying back shares.
Got it. This is all super helpful. Thank you so much.
Well, I'm glad.
For sharing with us today.
Thanks for hosting.
Thanks, everyone, for coming.