Equifax Inc. (EFX) Earnings Call Transcript & Summary

Hello. Hi. Good afternoon, everyone. Thanks for joining us today at our 41st Strategic Decisions Conference. My name is Kelsey Zhu, I'm the financial information services analyst at Autonomous. With me on stage today, we have Mr. Mark Begor, who is the CEO of Equifax; and Mr. John Gamble, who is the CFO of Equifax. Thanks for joining us at our conference again.

Thanks for having us.

Thank you.

So given the environment we're in, I thought it's only appropriate that we start with a macro question. And obviously, we've seen a lot of mixed data points on the state of U.S. consumers and the state of U.S. consumer credit markets. We have consumer confidence that's falling off a cliff. We have the Fed Senior Loan Officer Survey which points to some tightening around credit supply and credit demand. So just curious to get your thoughts on how healthy the U.S. consumers are, what are you hearing from lenders? And in that context, maybe also talk about what you're seeing for subprime consumers specifically?

That was a loaded question. So I'll parse some of it and I'll ask John to jump in, too. It's certainly an uncertain time with what's happening in Washington. I think that's the biggest change that's happened in the last number of months is the whole tariff conversation, what it's going to do to inflation potentially, what it's done to consumer confidence, as you point out. Those are all not positives for us. The positives for the consumer is they're still working. Unemployment is very low. And when you think about financial services, our customers being banks, fintechs, insurance, telcos, they're still going to do their lending as long as they feel like they get repaid and as long as they get repaid by consumers that are working. So consumers having the low unemployment is a good thing. That's one of the first indicators. When you see unemployment tick up, you generally see some of the origination activity slow down. You see cutoff scores change. That hasn't happened yet. So that's broadly a positive. Inflation is a challenge, particularly, as you point out, with subprime consumers, I'll come back to, because of where pricing is. And we really don't know what's going to happen with inflation around tariffs. Where is that going to land? Is that going to result in price increases that are going to pressure particularly the lower-end consumer going forward? So that's kind of in the future. The second thing that I think is important is our customers, being broadly financial institutions, are quite strong. The banks have strong balance sheets. The regulations that were put in place at the global financial crisis have really kept them strong. So they're still operating, I would call it, broadly quite normally, so really no change in that. There was a change in the fintechs, which are principally lending -- a lot of the fintechs are lending to subprime. That changed 2 years ago coming out of COVID when inflation went up. That pressured that demographic and there was a delinquency increase in subprime. Our view is that's pretty much normalized, meaning they were decreasing originations in really '22 and '23. As we got into '24, they've really more normalized, and that's as we go into 2025. You saw our first quarter results, and I want to come back to mortgage because there's a different dynamic in mortgage than there is in the rest of the industry because of what's happened to rates. But our first quarter results were quite strong. We beat our guidance. We saw what I would characterize as broadly normal expectation activity across the industry from a market standpoint and just some outperformance of Equifax, execution post-cloud, which I'm sure we'll touch on. So broadly, outside of what could happen with tariffs being implemented and prices going up, inflation going up, that hasn't happened yet. That's more in the future. We still see a fairly normal environment for us, which we think is good. And again, it goes back to consumers are still working, unemployment hasn't gone up. That part is quite strong. Mortgage has been quite challenging. This is not new. I think you know with the rate increases over the last 3 years, that's had a dramatic impact on our business as well as our competitors. Our mortgage activity is down in the neighborhood of 50% from what we would characterize as normal kind of pre-COVID, 2015 to '19 environments. And as we came into 2025, in kind of the October, November time frame, we thought 2025 would kind of be flat on a year-over-year basis. And then as some of the tariff conversations started late in the year and interest rates moved up, both the 10-year and the mortgage rates, we saw a decline in mortgage market activity. And that's when we guided for the year in the first quarter of mortgage activity. Think transactions, either refis or purchase transactions in mortgage being down 12% for the year. We did see in March a bit of an uptick when rates came down before Liberation Day. There was a bit of an uptick in refis and we ended up being down 9% in mortgage inquiries, which we define as market in the first quarter. But we held that minus 12% guide for the year, and that's a big impact on us. That's like $100 million of revenue, that minus 12%, that's in our guidance for the year. But when you think about the economy and the market, I think it's important to bring in what's happening in the mortgage market, which is heavily impacted by the doubling in rates over the last 3 years from historically low rates. And what's happening in the rest of the market, which I would characterize as broadly normal. We'll probably also touch on where this conversation around the market is principally on FI. I'm sure we'll talk about Workforce Solutions, where we have a large business that is growing in the government social service space, where we sell income and employment data to federal and state agencies for social service delivery. So that is a business which is a different customer set. In this economic environment, there's a lot of people that are still -- need social services. So that's got a market macro that's quite positive for us. And then the other big business that we're in that's outside of FI is the Talent business, we call it but selling data into background screeners. And while that's been impacted by slight declines in the hiring market, our performance has been quite strong as we've been growing new products and penetration. Anything you'd add, John?

It was quite complete. But in terms of delinquencies, I think as Mark said, what we're seeing is really concentrated in subprime. What we haven't really seen yet is where there'll be an impact related to student loans and student loan collections beginning and that's now starting to occur. But again, I think the place where you're starting to see delinquencies show up in student loans, again, is principally subprime, a little bit in near-prime. So it could cause for the subprime delinquency issues that we've seen historically to be a little larger. But it hasn't been that substantial as of yet, and that's something we'll be watching over the next several months. Again, as Mark said though, if you think about our large customers, our large customers generally lend into prime and near-prime. They're not really substantially impacted by subprime. So it hasn't really impacted our business in a very material way.

Got it. Super helpful color. And just to follow up on the macro question. You guys gave us a lot of helpful colors in the last quarter around how Equifax is expected to perform in a recessionary environment. Just want to follow up on that and get your thoughts around how different components of Equifax will perform in a stagflation.

Yes, that's one we didn't model. We haven't really experienced, as an economy, a stagflation environment, I think, from the '70s. What we did model for our investors in our earnings call was an update to our recession scenario. And we think about a more normal recession scenario, which is really what most economists were predicting. I think it's less so now. Who knows what's going to happen next week and how they think about what the tariff impact may be? But we looked at a reduction in GDP and then likely a cut in interest rates to boost economic activity in the back half of that recession. And I think what's important is around the changing mix of Equifax businesses. That's really what I hope investors saw in that analysis. We compared 2008, 2009, which is the last real economic event that we had. We updated it in 2021 and then we did it again for our April earnings call. And what you're seeing is a really change in the mix of our businesses to more and more of our markets and our businesses are recession-resilient or recession growth. And use the example of government social services. In an economic event, there's more individuals in the United States going after social services. That's a macro that's positive for that business. Identity and Fraud is another one. And I think in 2001, John, help me, it was -- 47% of our revenue was recession growth, meaning would grow in a recession. That was our model we did, call it, 4 years ago. When we did it in April, it was the high 60s percent. So really changing mix, more subscription revenue, which obviously is recession-resilient because it's kind of locked in during that economic period and a changing mix of the businesses. Anything you'd add around the changing mix?

No, obviously, the big growth drivers in terms of change in mix is government and talent. Government, obviously, very recession-resilient. Talent, actually recession-resilient to a degree because a fair number of our Talent customers buy on more of a subscription basis. It isn't subscription for a specific product but it's subscription for total value of transactions they'll execute with us. So given that, that's the case, more and more of our revenue is protected on the downside. It might limit growth, obviously, to the extent that you see a weakening economy, but we have a floor with many of those customers on the level of revenue that we'll achieve. So we're seeing a nice change in our mix toward more resiliency during a recession as we go forward. And as Mark said, given the fact that we're looking at a mortgage market that is down 50% from its current -- from the long-term run rate levels, we feel like there's a floor on mortgage as well, where to the extent that we see any type of interest rate reduction, that we're going to see some type of recovery in mortgage, which again is beneficial to us in a recession.

Just to clarify, Talent revenues are now mostly on subscription models, meaning it's still volume-based but there's a floor and ceiling of that?

Yes. Same with government. More and more of our contracts are going subscription. Actually, we have an Investor Day coming up in about a month. And one of the things we'll likely talk about is kind of that changing mix of our revenues from transaction to more subscription, which has been happening commercially and also strategically by Equifax because we think it's a good thing over the long term to have that.

Got it.

Talent wouldn't be most but Talent is meaningful, right? So I wouldn't say it's more than half, but it's meaningful that we have some significant customers that buy from us on more of a subscription basis.

Got it. And I love that Investor Day preview. Please keep them coming. So outside of macro, I would say the one question that's probably top of mind for a lot of investors is the recent comments made by the FHFA Director Pulte. So just curious to get your thoughts around kind of how you're thinking about the ceiling of mortgage credit files.

Yes, so it's a complicated question. I think everyone in the room knows that the new director was at an MBA meeting and had some comments in a Q&A session about FICO pricing. Issued a tweet, I think, last Monday -- last Tuesday and then another one on Monday that likely were around FICO pricing. So I haven't met with him. We haven't met with him. He's very new in his position so we're not really sure what his direction is. We're focused on what Equifax can control and what we -- how we go to market. The ceiling on the price of the credit file and the FICO score. As you know, it's a mandatory -- it's mandatory 3B and it's also a mandatory FICO score in the current environment or the current regulations. I personally don't expect that to change because there's so much history with the FICO score and there's so much value in a 3B pull. On the credit files while we're very similar between TU, Experian, and Equifax credit files, there's some meaningful differences in the contributors, which is why a 3B is very valuable. There's 10 million U.S. consumers that are only on one credit bureau. Think about your local bank. You only have a credit card and auto loan with a local bank. They only contribute to one of the three credit bureaus. Not all 30,000 financial institutions contribute to all three of us. So that creates differences, which is why in a big-ticket transaction, there's generally more data use, which we think will continue. There's 40 million consumers in the U.S. that have a 40 basis -- 40-point difference in their credit scores between the three bureaus, same thing because of the contributors. So there's a lot of value in pulling more than 1 and pulling 3. Some of the most sophisticated lenders in the United States pull 3 when they're not required to because they see the value of getting additional data. So I think that's one that I wouldn't see changing inside of the mortgage environment. It's a big-ticket transaction. The federally guaranteed mortgages are focused on promoting housing. So the idea of excluding someone if they're only on 1 credit file doesn't seem like it's aligned with a change going forward.

I guess in that context, are you still expecting the implementation from tri-merge to bi-merge, single score to...

No. But everything we know is the idea of sort of what everyone may remember, the prior FHFA was studying and they never concluded the study. They actually suspended the study, I think early this year. They were studying the idea of FICO and Vantage scores being pulled in every mortgage and relaxing the 3B requirement to 2B. The industry pushback on the 3B to 2B for all the reasons I described is that there's just so much more data in a 3B pull than there is in a 1 or 2B pull. So that was 1 even if it went through, our expectation was most, or all lenders would still pull 3. Everything we know about the new administration is there's no discussion around 3B, 2B because that was kind of put to bed by the industry over the last couple of years and the different comment periods and studies that took place. I think the issue now is the new director's focus on FICO pricing.

Got it. And speaking of mortgage prequalification, which is like the one market where lenders are not mandated to pull all three files, maybe talk about the trends you're seeing there. Are most lenders sticking with 3B? And I know Equifax recently launched a new product that combines income employment data and credit data on one single file. So maybe just tell us more about the pricing of that product. What exactly is included and the market share that [ you're ] seeing?

Yes, sure. So what Kelsey is referring to is that in the shopping process for mortgage, there's not a requirement to pull all three. And for cost reasons, this is what we understand from the industry, some of the mortgage originators have gone to either a 2B or a 1B pull during the shopping process. Remember the shopping process, they've got a funnel of consumers that are coming into their digital websites inquiring about a mortgage, and they're trying to make decisions on who do I spend time on. I don't want to spend time or COGS or people costs on someone that I can't close or get to application. And because of cost reasons, they've gone to either a 1B or a 2B pull. We launched a new solution a couple of months ago, where we're going to add an income and employment or we call it TWN, TWN is the acronym for the work number, indicator on our mortgage credit file to try to differentiate Equifax's credit file from our competitors, TU and Experian's. We think we can add a lot of value because as you know, in a mortgage process, they verify credit as a part of the application on a 3B basis, and then they also verify income and employment at closing, meaning you have to be working, you have to verify what the income and employment is. By giving an indicator upfront, which during the shopping process, by just pulling a credit file, you're really blind to is Mark working. It might be Mark with a decent credit score, but I don't know if he's really working. So we want to differentiate our credit file upfront, which is why we rolled out this TWN indicator for mortgage. We're doing the same thing in auto, in personal loans. We're further along in auto. P loan will come next, and we're also going to do something in card. And this is really all post-cloud. As everyone in the room probably knows, we've been investing over the last 5 years to move our data and technology into the cloud. It was a huge project, a huge investment. This is a great example of the kind of things we can do that I've wanted to do for years as the CEO, which is really to leverage both our credit data from USIS with our income and employment data from EWS. So we're really trying to position our credit file as one that is going to provide more value. We're not going to change the price for our credit file, meaning we're going to deliver our credit file with more value in it by having that TWN indicator around. Mark's working, might be a TWN indicator with Mark works for the company, Equifax in my case. We're also testing some kind of income band like Mark's income last year was X. And that gives the mortgage originator more value in that shopping process in order to prioritize which consumers that are in that digital funnel that they're going to work on and we think will drive market share for our credit file. So that's really the approach that we have. We're just in the early days. There's a lot of workflow changes that will have to happen in order to be able to use that effectively in that shopping process. But we think by pricing it at the same price as our credit file and just providing additional value, that will encourage mortgage originators to work that into their workflows, and we're seeing positive feedback. From everyone we talk to say, "Yes, that adds value." And I think you know that I was a lender prior to Equifax so I know the value of credit and I know the value of ability to pay, which is income. And having those 2 together really provides a lot of value. As I said, we're going to do the same thing in auto. Auto loans in the subprime, near-prime verify credit and income and employment at closing, having that visibility when you're an auto lender or at a dealer around Mark's credit score, which they always pull upfront in the shopping process in auto, generally a 1B pull in auto, understanding that Mark's working, because they have to verify income and employment, gives them more confidence. Now understanding Mark's income is in this band helps them think about which kind of car should I be prioritizing, putting in front of Mark in that shopping process? Same thing with P loans. They verify income and employment. It's a large unsecured loan, big ticket along with credit. So if we can help them early on to prioritize which consumers do I have more confidence in closing on because I know their credit score is X and they're working, where their income is. And obviously, the balance we want to have is deliver value in that shopping process that differentiates our credit file and drives more market share but still protect the full TWN pull that happens in mortgage, auto, P loan. And then in card, we're going to do the same thing, a TWN indicator on our card credit file or credit score that say that credit score is X, and Mark's working or an income band, we think it'd be very valuable inside of the credit card space. And these are an example of kind of our post-cloud focus on innovation and really leveraging the broad set of assets that we have across Equifax. Beyond the TWN indicator on the shopping credit file, we're also adding our cell phone utility attributes. And we have a large data set we call NC+, NCTUE is the acronym. It's a data set of cell phone utility payment records that are not in the credit file. It's on 220 million Americans so a lot of coverage there. For the mortgage file, we're adding some of those TWN attributes, another example of adding more data to our credit file to differentiate against our competitors to really drive market share.

I think it's brilliant to combine USIS and EWS data. I guess my -- what I was kind of wondering is if pricing stay largely consistent, how do you think about the cannibalization risk for [ TWN? ]

Yes, we have to balance that. It's definitely something that we're laser-focused on -- of adding enough value to make our credit file more valuable upfront in shopping while still protecting the full income and verification. Remember, when we do an income and employment verification, we deliver 50 attributes. Generally, I think, as John pointed out, it's a trended data asset. We'll sell 3 years' worth of data because the mortgage originator wants to make sure that Mark's incentive income is consistent. We have 50 different attributes. So a lot of data is delivered in that closing income and employment verification. So adding that right balance around enough upfront to differentiate our credit file to be more than credit and add some visibility around Mark's working, Mark's employer name, maybe an income band, but still protect the full pull at the end. And I think that's a balance we can work through from -- and the feedback we're getting from the mortgage industry and the auto industry that we've started dialogues with is quite positive. And again, baseline is we're not going out to sell our credit file at a premium with this additional data. We're going to sell it with the additional data at the same price.

And we think NCTUE actually expands the pool of available borrowers. So it actually increases the market size because borrowers that are on the margin, adding the NCTUE data can make them eligible for a mortgage, which increases not only their opportunity but also our ability to sell.

But this is what you would expect us to do at Equifax. We believe we have scaled data assets that are advantaged versus our competitors. I think it's hard to argue TWN isn't one of them, the income and employment data. But NCTUE, we have other alternative data. But with that now our post-cloud environment where we put all that data together in a Single Data Fabric that's keyed and linked, we can now surface these kind of products and solutions with our focus on innovation of new products quite quickly. So it's something that we're super energized around the market opportunity for us going forward of leveraging the broad array of Equifax assets. And as you point out, we've got a big business in our U.S. credit file business. We have a big business in our U.S. income and employment business we call Workforce Solutions. Leveraging those two is a big opportunity. We made an org change in January of this year. We used to have two mortgage leaders, one for each business that would sell into the mortgage industry. We now have one leader that owns both solutions and you can see what's happening. The idea of putting these products together post-cloud is really a big deal.

And I'm sure we'll talk about it. But as we're now completing the tech transformation, and we now are cloud native on GCP, what you're finding is we're able to deploy very high-performance analytical assets and decisioning assets that our customers can use with this combined data set to help them build products faster that we can deploy faster. So it's not just a matter of selling them individual point data solutions. We can now sell them combined data solutions that they can use either as a combined score or as something in their own decisioning workflow where they determine where to use the data assets more easily and quickly and they can deploy on our infrastructure. So these things are happening quickly not just in the U.S. but actually with the analytical and decisioning assets that's happening around the world.

This is all super helpful. Maybe just one more question on USIS/consumer credit trends. I think especially in segments like auto, we've seen some demand pull forward in March and April trends. Maybe just talk about what you're seeing in May across auto, car, P loans and other categories.

Yes. I think we'll probably go through our guidance that we gave in April, but we wouldn't expect that to change as we go forward. There was a little bit of tariff pull-through on auto but not real change in the others. I wouldn't call it meaningful. I think as you know, in the House Reconciliation Bill, there's a proposal that auto interest would be deductible, which obviously would change perhaps auto behavior pull forward in the future, some auto activity. But I don't think we've seen any real change. John?

I think the last thing we said is our guidance so that's where we are.

Got it, got it. Let's switch gears to talk about EWS. I feel like there's a lot to discuss here, and maybe let's start with the competitive landscape update. Obviously, we've seen a lot of news flow in this space, whether it's like some fintech players announced that they've been gaining inroads with government agencies, whether it's like Checkr acquiring Truework. So maybe just give us an update on what you're seeing as the competitive landscape evolves and how that influence the way you think about pricing.

Yes, so we think we still have a very strong position in EWS. We have just real scale in the business. At the end of the first quarter, we ended with 4.4 million companies delivering data to us every pay period. There's different numbers out there, 10 million, 11 million, 12 million, 15 million companies in the U.S., but we got a lot of them. We ended with 138 million individuals in our data set every pay period. There's roughly 220 million, 230 million income-producing Americans, so we've got the 90 million to go. In the first quarter, we grew our TWN data assets by 11%, so strong growth and the ability to add assets. We've got a dedicated team focused on that. So we think we've got a strong competitive position. We added three new partners. I think you know we get records both directly from companies where we deliver regulated services like I-9, Work Opportunity Tax Credit, Unemployment Claims management, W-2 management, other solutions like that. And then we do income and employment for free for that company, so the HR manager outsources that to us with the other activities. That's about half of our records, and then the other half come through with partners. And that scale as well as our scale of distribution is quite large. Some of the competitors you mentioned are more consumer-consented activities. There's a lot of friction with that. And we have consumer-consented solutions. We're actually investing in some new ones like for the government space, we want to have a consumer-consented solution primarily for the gig government social service recipients if they're doing self-employed type income. But there's so much friction in that. We haven't seen that as being a big competitor to Equifax around income and employment data. With regards to our pricing in income and employment with the TWN, we think we're very balanced on pricing. We take price up every year in all of Equifax businesses. We're quite balanced about the price around our value and want to make sure that we're being reasonable on the price. Nobody likes a price increase in any business. But price is only one of our growth levers in EWS. We already talked about record additions. Remember, when we add new records every quarter, that drives our hit rates because we're getting every transaction from our customers. Today, we fulfill for the records we have. So when we grow our records 11%, that drives our revenue growth up for the different verticals that we participate in. We have the ability to roll out new products. That's a big part of our growth engine in EWS is roll out new solutions that really either have more trended data or other data combinations. And as you know, we've been adding a lot of new solutions, new data elements, whether it's incarceration data from our Appriss Insights acquisition that's used in Talent and government, we're really the only data set there. We've got an exclusive partnership with National Student Clearinghouse that brings education data in for talent. And we're continually looking for more data assets that really drive that. But then the big lever for our EWS business is penetration. Our biggest competitor when we think about competition in the EWS space is manual verifications that are done in every vertical. Mortgage, if we don't have the records that mortgage underwriter is doing a manual verification, same thing in auto and P loan. And then in background screening for Talent, our vertical we call Talent or government social services, those are huge TAMs. The government TAM is about $5 billion of total verifications done. We're about $750 million. So there's another $4-plus billion of manual verifications at the federal and state level for delivery of government social services. We're growing into that conversion of manual into our instant solution. Same thing with background screening is the background screening is very much a manual process for a lot of the background screeners. That's about a $400 million business in a $4 billion TAM, so lots of penetration opportunities. We don't have the big penetration opportunities in other businesses like we do in EWS, where our principal competitor in EWS is really the manual verification efforts that companies do on their own.

And Mark already mentioned trended, right, but it's important to realize we have over 750 million total records. And we have a record -- our penetration on the number of U.S. citizens where we have at least some record is extremely high, not quite 100% but pretty close. So we can provide a response on a trended request a very high percentage of the time, and which provides a very strong competitive position for us relative to any other solution because of the fact that we can provide not just current information but also a lot of historic information. And if you think about the level of revenue we generate from trended information is on the order of 50%, a little under 50% of the transactions we execute are for trended information, meaning current plus historical.

And maybe just one -- and maybe you'll get to it, Kelsey, but just as a reminder, EWS -- so our long-term growth rate for Equifax is to grow 7% to 10% organic, 1 to 2 points of rev growth or 8% to 12% for overall Equifax. So 7% to 10% organic growth, we expect EWS to grow 13% to 15%; USIS, 6% to 8%; International, 7% to 9%. So it's highly accretive to our growth rate, EWS is. And I think as you all know, EWS margins are 50%, which is very accretive to our kind of mid-30s growth rates -- margin rates. And really, a big part of that 50 basis points of margin expansion we get -- we expect to get over the long term in our long-term growth framework.

Got it, and that's super helpful. I do want to just follow up quickly on pricing. I feel like it's been a while since we talked about pricing's contribution to growth across mortgage, Talent and government in 2025 and also how that compares to historical levels.

Yes. So we don't talk about pricing for competitive reasons. But I can tell you, if you think about our long-term growth framework for EWS, and I'll use that as an example, you could use -- I could give you the one for Equifax. But we expect EWS to grow 13% to 15% over the long term. And if you think about the building blocks of that revenue growth, over the long term, we expect a couple of points of GDP. So think a couple of hundred basis points of that 13% to 15% is just going to be broad economic growth. We would expect record additions to be 3 or 4 points of that revenue growth, adding records every year. Higher hit rates drives our revenue. We would expect our product rollouts to add 3 to 4 points of revenue growth annually from new products, whether it's our trended solutions, data combinations, all the new products we're rolling out. And then penetration is another 300 to 400 basis points. That's a big part of our growth algorithm is that big TAM we have in government, big TAM in Talent. We have penetration opportunities in mortgage, in auto, cards, and P loan. And then price is the last piece and that's a couple of hundred basis points. So if you look at those building blocks, you see a very balanced element of multiple levers to drive growth. It's like four legs on the stool of records, product, penetration and price. And those to us are quite balanced when we think going forward. In some years, one might be higher than the other. Product has been a big piece for the last couple of years. Records have been a big piece for the last couple of years. But the fact that we have those multiple levers, we think is a very balanced kind of growth algorithm that gives us a lot of confidence in that 13% to 15% over the long term.

Got it. Maybe just one quick follow-up on record growth. I think you've talked about expanding into 1099 [ patient ] records. Maybe just talk us through your strategy there as well as how many incremental records are we expecting for '25 in the medium term?

Yes. Some of those we're not giving guidance on like incremental records in '25. So when you think about the working population, call it, the 225 million income-producing, I said working, I meant income-producing, that's made up of roughly 100 -- nonfarm payroll is like almost 170 million, I think it's 167 million W-2 nonfarm payroll participants in the United States. Then there's another 30 million or 40 million self-employed. That includes doctors, dentists, lawyers, architects, accountants that have their own businesses, that have 1099 income. It also includes a large number of gig workers. So think about Uber drivers, DoorDash, self -- somebody who works for a landscape company that pays them on a 1099 basis. So a large population of 1099. And then there's another 25 million, 30 million individuals that defined benefit pensioners, that's income. And a pensioner might get income at age 55 as a fireman or a police officer in a state or federal or a local agency and then get a second job. But that pension becomes part of their income. So we've got a dedicated team that are focused on companies. That's the 4.4 million companies that we have getting those records. Those are principally W-2 records, principally nonfarm payroll. There's some 1099 that come with those. We've got a lot of penetration. There's a long way to go. So just as a reminder, we have 138 million individuals against the, call it, 170 million nonfarm payroll plus the gig plus the pension. We also have another dedicated team on the 1099. Different strategies there. Some of those we get from payroll processors, some we get from companies that pay individuals to be 1099 for them. So that's one way to get those records. Some of the payroll processors have operations to do self-employed individuals. There's tax preparation services that's another avenue to get those records because you have to do an estimated tax payment every quarter. So that's a way to get those records. And then on pension, there's companies that have defined benefit pensions. Think about legacy big companies in the U.S. that had historical pensions and no longer have, they're still making those payments. So we go to those companies and say, "We'll do income and employment verification of your pensioners, just like your employees for free." And they outsource that activity to us. There's some -- think about like a payroll processor but a pension administrator, Fidelity, State Street, other big companies that do pension administration. Those are targets for us to go in and have that same kind of partnership with them. And then there's a whole bunch of pension records with federal, state and local agencies like the Fire Department of New York, universities, education systems, state employees, local employees. We have dedicated teams focused on getting those. So as I said, we have a dedicated leader that's responsible for record additions in that whole remit that I talked about that works directly for the leader of EWS who works for me. So it's a big job. They're dedicated on it. They've got people assigned to each of these different verticals to go after the records. So we have a lot of confidence in our ability to grow records going forward. We don't give guidance on records. We give guidance on probably too many things. But we talked about the long-term framework of getting 2, 3, 4 points of revenue growth from record additions. And there's just a long runway when you think about 138 million versus 225 million to 230 million total available individuals out there. And as I said earlier, we have -- for W-2 records, we really have two avenues, going direct to companies with our Employer Services business where we deliver those regulated services like I-9, Unemployment Claims and the other solutions we deliver, and we get employment records in return, and we do income and employment for free. And then we also do through partnerships. In the first quarter, we added three new partners. Last year, we added 15 where we're taking their records in. In the last 4 years, I think we've added 50 partners. So we've had a good track record of going after partnership records, direct records, 1099 and pension.

Let's talk about the government vertical for a second. Has DOGE been more of a tailwind or a headwind for Equifax? I feel like you can make the argument either way if you're...

Yes, we would only make it that it's -- I would broaden it from DOGE. I think the current -- our current President, I think the Republican Senate and House, I think the kind of administrators he's putting in place, there's a big focus around the deficit. There's a big focus around spending. And there's a big focus, in our case, the world that we live in around social service delivery and the improper payments. And I think there's broad alignment both Republican and Democrat, to deliver social services, have that safety net for those that need it and those that deserve it and qualify. The federal government, this is not an Equifax number, quantifies the improper payments of $160 billion per year. That's a big number. And what that means is $160 billion of payments to people that probably qualified at one point but no longer qualified today under the requirements for whether it's Medicaid, Medicare, food stamps, rent support, childcare support, earned income tax credit, all the different social services, that safety net that's in place there. And there's clearly, in the last, I don't know, 120 days, meaning since President Trump got in office, been a different tide in Washington. I've been down there three times since the inauguration, meeting with agency heads, meeting with OMB, meeting with other administrators about how can we help with our solution to attack that $160 billion. And there's just a big focus on that, which we think is going to be beneficial to Equifax. And part of it is DOGE focusing on how you can you improve spending in Washington. We're not in that category of companies that they're looking to cut costs on. We believe we're in that category of companies that can help us deliver solution against the $160 billion of improper payments. And if you think about our government TAM, we think about a $5 billion TAM if our product using income verification was used universally across all government social services versus our $750 million of revenue today. So think about a $5 billion cost at full implementation, which is going to take a long time, but that's against $160 billion of improper payments. There's a huge ROI there with our solution. So as I meet with the agency heads, there's a real focus on how do they address improper payments. And you've seen in the House bill that was approved a week ago, you've seen in the articles, the same ones I've read or if you've read the bill yourself, which is quite inclusive, there's a lot of elements in there around strengthening the income requirements for the delivery of social services. There's a work requirement now for Medicaid for certain individuals that wasn't in place before. That's one we can deliver on because one of our attributes is hours worked. So we can help deliver to those that deserve it, the social service that those shouldn't be getting it and make sure that they're not getting it. They're changing. Right now, there's a redetermination. If you get social services today, there's a requirement 12 months from now to check again to see if you qualify. In the current House bill, there's a move to make that every 6 months. That's going to be more frequent redetermination. So there's a real focus in Washington around how do we address the improper payment side, which is really our focus inside of the business. And then when you think about the individuals that are getting social services, it's a very complex demographic. They have lots of life events. They come in and out of the workforce. They have a parent or a child and they either can work or can't work. And there's just a lot of variability in that. And that's why more frequent redeterminations, we think, are quite positive. The other thing that happened in the prior administration is the relaxing of some of the standards around income verifications. There was waivers issued during COVID that were extended all the way through into 2025 to relax the income standards around income verification. We think that's going to be rolled back. So that's a broad tide that we think is positive for Equifax. When you get -- when you think about the TAM, there's a large piece of the TAM at the federal government, think 20%, 25%, but most of it is states because in the U.S., social services are delivered by the states. They're the ones who do the adjudication, the verification of eligibility of their residents, the people that are coming in to get those social services. And then we talked about on the earnings call a few weeks ago in our April earnings call that we had a very large authorization that happened during late March with a Social Security Administration contract for disability income of over $50 million. And I think that's a proof point about our data is really valuable in this case, around improper payments. People that are on disability income but then get a job and no longer qualify or get an increase in their income and no longer qualify for those disability payments. Our data is now going to be used to really identify those to really take them off of the disability when they don't deserve it. So we view this as a real tailwind.

Mark, you mentioned the SSA contract, and I know historically, there's been some ups and downs with that contract, sometimes not fulfilling the growth expectations that many people may have had. So maybe just talk us through your confidence in achieving that revenue target for the SSA contract.

Yes, I would just say maybe stepping back broadly, look, doing business with the government is complex. It's complex bureaucracies, it's complex environments. There's multiple participants. In our case, you've got kind of the federal agencies are setting the standards that the states have to follow but the states implement it. So that level, there's 50 states. Each agency is generally a different organization in each state. So that complexity of interacting there. And then there can be changes in administration's focus. You saw in the last administration, the relaxing of some of the standards that have historically been in place by, in my opinion, the prior two administrations, Obama and Trump 1.0, had fairly strong standards around income verification. Again, the concept is everyone wants to have a safety net to those that need it and qualify for it. Those that don't, there's a view that they shouldn't be getting those. That was relaxed in the last administration and SSA was probably an impact of that where there were some changes there. We think going forward as they codify into law, things like redeterminations being done twice a year, those things are really positive for us over the long term. And I think broadly, there's bipartisan support around get social services to those that deserve it. Those that don't -- that are either fraudulent or improper payments, those payments shouldn't be made.

And with government, it's important to remember, it strongly benefited by record additions, right? So it's -- you're looking at current income validation as we add 10% or 12% new records in 2024, and the progress we're expecting to make this year, that's directly beneficial to the government business. It obviously helps us in what Mark's talking about in terms of ensuring proper payments are made. It also helps us in terms of boarding people into benefits faster so that people get those benefits quicker, and they can get the benefit of them on a much more rapid basis. So I think it helps both sides of the equation and record additions is a huge driver of the growth in our government business.

I think the coverage we have is just so massive. I mentioned 138 million individuals in our data set every pay period at the end of the first quarter but 198 million total jobs. So what it says is that in our data set at the end of March, we had 60 million people that had two jobs. And you think about that, that's incredibly -- these are two W-2 jobs. So they're working 20 hours at one nonfarm payroll, think about Walmart, JCPenney, Lowe's, a restaurant, a warehouse, et cetera, getting a W-2 income and then a second one getting a W-2 income. And one of the things we hear from our government customers is the complexity of that consumer base. They'll generally have a W-2 job, maybe two that are part time. But then they also have a gig work. They have a gig job, where they're driving for Uber or DoorDash or something like that on the weekends or Lyft. And that income is very hard to collect. And one of the things we're rolling out in the second half of this year for government is an integrated solution where the government agencies will be able to hit our database and look at the records we have in our database, think about the 138 million. And then if we don't have the complete picture of that individual, a consumer-consented solution where we'll pull bank transaction data that's integrated in our workflow in order to add that bank transaction data and which will pick up those 1099 jobs. It might be an Uber job, a DoorDash job, a Lyft job, pick up that income so we can have the more complete picture on the consumer. So an investment in technology and product, when you think about our focus about priorities for where are we investing in people, where we're investing in new products and technology and bolt-on M&A, EWS is at the top of the list for obvious reasons. It's our fastest-growing business. It's our highest-margin business. We're continuing to make big investments there to keep that engine going.

Mark, why did -- in your view, why did the CMS change their funding practice last year from funding 100% of data cost for the states to...

Yes. There was -- and just a little more background on that. For a decade, CMS, which as you know, is Medicaid, Medicare had a contract with Equifax for 10 years, where the data cost was paid for by the federal government and the states could access that data for free because they wanted to promote -- they're on the hook for most of the social service costs so they want to promote the use of verified data. So they said we'll pay for 100% of those costs. The prior administration last summer moved it to 75-25. Unclear to us about why. Budget-driven, we're not sure. That obviously had impact on some of the states that didn't have budgets to pick up the 25%. That was something that we are recommending that the new administration put back in place. The 100% reimbursement, we would expect they would. It makes sense to do that. The same change also happened with USDA. USDA had 100% reimbursement in the case of food stamps, SNAP food stamps, those are 100% paid by the federal government but delivered at the state level under their guidelines for income verification, USDA put in a program that looked like CMS a couple of years ago, and then they went to 0 reimbursement in 2020 for -- we are encouraging them to put that back in place. So verified income is used at the state level.

Got it. I think we're running out of time. Maybe just one last question on capital allocation. You talked about the $3 billion share buyback program, along with the 20% to 30% dividend payout ratio. Maybe tell us more about your M&A strategy here, key products, data asset, key region you want to expand into?

Yes. So it was a long time coming. We've been talking about doing this for quite some time since we launched the cloud investment. But we knew we wanted to put a formal long-term capital allocation plan in along with our long-term growth model. Long-term growth model is 8% to 12%, 1 to 2 points of M&A, 7% to 10% organic and 50 basis points of margin expansion. We've added to that a cash conversion framework of 95% plus of our earnings, converting into free cash flow, which is very, very high. And then we rolled out in April a capital allocation plan where we're going to continue to invest in Equifax. That's clearly our first priority with our high growth rates and high margins. We want to keep investing to deliver those kind of growth rates and margin expansion. We'll continue to invest in CapEx at about 6% to 7% of revenue, which is about $0.5 billion per year. That CapEx going forward will be principally around growth now that the cloud is complete, we have a new tech stack. So that's going to be around driving new products, driving innovation. We have a long-term framework of 1 to 2 points of rev growth around M&A. That translates into $500 million to $700 million a year of tuck-in and bolt-on M&A. No big deals. We have 0 interest in that. We want to strengthen the core of Equifax. Four strategic priorities: Strength in EWS, about half of our 20 acquisitions, at least the TEV over the last 5 years has been in EWS. So strength in EWS. Differentiated data, and we've done a number of acquisitions there. Appriss Insights is a great example, PayNet, DataX, Teletrack, so differentiated data with a moat around it. Identity and Fraud, big macro. And then international platforms like Boa Vista in Brazil. So those are the kind of the areas we want to do with M&A, very disciplined around the financial criteria. It has to be accretive to our revenue growth rate and into our margins. And so if you look at that, that's roughly $1 billion-plus per year investing in Equifax growth between growth CapEx and bolt-on M&A. And then the remaining free cash flow we want to return to shareholders. We laid out a dividend increased it 28% in May for 2025, but going forward, increasing it in line with earnings with the payout ratio of 20% to 30% that you described. And then significant excess free cash flow in the buyback. And we laid out a 3-year -- I'm sorry, a 4-year $3 billion buyback going forward. One last point I'll make on that is that we talk about the mortgage market recovery. And if and when interest rates come down, if and when mortgage market comes back, that will all be accretive to the -- and likely end up in buyback. We're not going to invest more in CapEx. We're investing the right amounts. We're not going to invest more in M&A. We're going to be disciplined around that. Where that excess free cash flow will go in a mortgage market recovery is to a bigger buyback.

Got it. This is all super helpful. Thanks so much for sharing with us, Mark and John.

Thanks for having us.

Thanks a lot.

Thanks, everyone, for joining us.