All right. Good morning, everybody. Thank you for being here. Welcome to day three of our Financial Conference. For those of you who don't know me, my name is Manav Patnaik. I'm Barclays' Information Services analyst, and we're very pleased to have with us here again Equifax. And for those of you who don't know, we have Mark Begor, who's the CEO, John Gamble, CFO, and also in the audience down there, Trevor Burns, I'm sure many of you spoken to. So thank you all for being here.
Thanks for having us. Great to be back.
Mark, maybe just to start off with a kind of big-picture macro questions, and maybe we'll go by vertical. But, you know, we've had a lot of banks, card issuers, all your customers probably, who've had their own opinion on the health of the economy, et cetera. So maybe we can just start on the non-mortgage side-
Sure.
And just talk about where you see, you know, car, auto, the consumer, and how things are going there.
Yeah, I think broadly, the economy is still pretty strong. You know, I think you heard that this week, and we've seen it for quite some time. The recession that CNBC keeps talking about, although less so now, seems to be getting pushed out and maybe won't happen, at least from our perspective. And the way we look at it is, low unemployment, high number of people working is a big deal. If people are working, they can pay their bills, they can pay their credit card bills, their auto bills, they can take out new loans. So, you know, that is a real positive. And generally, unemployment is aligned with delinquency increases, and when delinquencies increase, our customers pull back. They get nervous about the future, and we just haven't seen that.
You've seen delinquency increases in subprime, and that's been happening for the past four quarters, maybe a little bit longer, and mostly impacted fintech. Most of our core customers don't do a lot in subprime. It's more on the fintech side. So, you know, broadly, the consumer is still pretty strong. You know, they carried over some savings from COVID that they still have balance sheets. Their credit scores are still above 2019, you know, which is a good sign. So when we look forward, that's a real positive. Our customers are also strong. You know, back in April, when we had the deposit scare, you know, with some of the banks, there was concerns that that was going to have a broader contagion. It hasn't, which is good news.
But our customers are really strong. And if we all scroll back to the last big economic event that we had of scale, largest ever, actually, in 2008, 2009, it was really the banks were in trouble and consumers. We really don't have either now. So that's, you know, a pretty good environment going forward. We put in our outlook for 2023, a slowdown in the second half, and I think we've seen that. We've got some pockets in non-mortgage, particularly in the talent sector, where, as you know, in Workforce Solutions, we sell our employment history data from the TWN database into background screeners. And we've seen really a bifurcation of the labor market, where blue collar is still quite strong. White collar really probably a year ago started tightening up.
You saw companies over the last year either doing layoffs or doing hiring freezes. You know, and those result in less background screens done, and it also impacts our employer business, where there's onboarding for our I-9 business. So, you know, that's one kind of isolated macro that you know, we've seen some impact outside of mortgage, which I'm sure we'll touch on. You know, but broadly, when we look you know, forward to 2024, you know, it feels like the consumer is going to stay pretty strong. You know, it's hard to see an economic event out a couple of quarters right now, given you know, how employment is still fairly robust. I forget the numbers. There's 9 million open jobs and 4 or 5 million people looking for jobs. That's a pretty good environment.
Got it. And so if I look at just USIS non-mortgage, you know, in a soft landing scenario, you know, what are some of the drivers that you guys can control? Because I think right now we've been run rating in, like, low single-digit growth, I think so.
Yeah.
You know, I think it's fair to say maybe we're not going to-- the consumer is not going to get much stronger, but let's just say they flatline. How should we think about-
Yep
the growth there?
Yeah, so USIS has its core credit business, which is more impacted by the consumer and what our customers' decisions are around underwriting. You know, what kind of new card originations, auto loans, you know, where their cutoff score is going to be. And with delinquencies being under control, we would expect that to be fairly normal, meaning we don't, we don't see, you know, a tightening there. And you know, that's a business that should grow, you know, in the kind of mid to low single digits, you know, longer term. And just as a reminder, for USIS, we think about USIS as being a 6%-8% revenue growth business for us inside of our 8%-12%, obviously, EWS north of that. So in that 6%-8%, you know, they should be at the lower end of that.
We have a couple of businesses that are outperforming that 6-8 quite strongly over the last couple of years, and we expect that to continue. One is our commercial data business. This is where we sell data to financial institutions, telcos, insurance companies for small businesses. And that's a combination of the entrepreneur's consumer data and then commercial data that we've collected over different verticals, including bank transaction data, leasing data and other data sets that we have. And we've got a fairly robust commercial data set. Last year, that business was up 20% in USIS, so way outperforming that 6%-8%. So far this year, they've been up double digits because of some of the new solutions that we rolled out from our cloud implementation, and really, they're differentiating data.
The second business in USIS that should outperform that 6%-8% is our identity and fraud business. And as you remember, we bought Kount a couple of years ago and then last year, Midigator, and those businesses are growing double digits inside of that digital macro, where more and more of our customers' transactions are going online... and when they go online, you have to verify the identity of the individual or small business every time they interact. You got to make sure it's not a fraudster coming in in that first transaction or when they come back again, and that's really been a strong growth for us. That market's growing kind of high teens, and we've been, you know, growing at the same pace.
We're really excited about the integration of Kount now as we put together all the Kount data. Kount, just as a reminder, brought to us very unique e-commerce identity transaction data. We had a pretty wide array of financial identity data from our credit file and some of the other datasets that we have of people paying their bills every month. We know that it's Manav, from you paying your bills to the bank, you know, for your credit card, your loan, your mortgage. Now we have e-commerce data that shows your shopping behavior. You know, so it shows you shopping multiple times per week, and the frequency of that Kount data is really quite powerful when it's combined with the bank transaction data we have.
We're starting to put that together, and we're seeing big lifts in the score, performance from an identity standpoint, which is really what's driving, you know, our revenue growth. So we like identity. It's one of our priorities around M&A, you know, and those are probably the two businesses that, you know, in USIS are outperforming. I'll make one last point on USIS. As you roll forward to 2024, they'll complete the cloud. They've got a couple quarters left to finish their cloud transformation. We believe USIS, like the rest of Equifax, is gonna be significantly benefited by being fully cloud native. Delivering those always-on stability. In a digital macro, when you're interacting with a financial institution, if you're a data provider, you got to make sure you never have any downtime.
We believe the only way to get nine nines or always-on stability is in the cloud. It's very hard to do in a legacy environment. Also, the latency or speed of data transmission is another thing that's got a lot of friction when you're in a digital macro. Those two, along with differentiated data and the product rollouts, we believe should advantage USIS in 2024, 2025, 2026 cloud native around picking up some share, moving from secondary to primary because we're gonna be always on. That's another positive when we look forward for USIS in 2024 and beyond.
Okay. I was gonna ask about the tech later, which I will, but maybe since you brought up the share gains point of it. So to date, you know, it's been four or five years since the breach.
Yeah.
You know, so I'm implying maybe from what you said, you're still in that secondary level and the opportunity to get back to the first. So how do you think about the share opportunity or loss even today?
Yeah. So just maybe pulling back, as you know, these market positions are quite sticky. When you get in a primary or secondary, and what we're talking about is most financial institutions will have a primary credit file data provider, typically by vertical. You know, it's generally not the, you know, institutional, although some are, and we're primary in a bunch of financial institutions, so are our competitors, right? But we're not primary everywhere. We didn't lose much after the cyber event in 2018, meaning we didn't go from primary to secondary. There was a... As you know, our revenue was only down a few points, because it's so sticky. When you get in a primary position, the data analytics team inside of that company gets used to using your data.
They have years of history of using the data, so they really understand it. So that switching from primary to secondary has to be a catalyst. There has to be a reason to do it. We think cloud is one of those that will drive that. And primary generally means you're getting 70%-80% of their credit files on business or scores, and the secondary is getting 20% or 30%. So where we're secondary, we want to move to primary. You know, and we believe the $1.5 billion investment we've made in the cloud, the always-on stability, our differentiated data, the new solutions we're bringing to market, which I'm sure we'll touch on from our new product initiatives, will make us a stronger partner and allow us to make some of those secondary to primary moves, which should drive USIS revenue.
We have active dialogues going around the power of the Equifax Cloud, you know, as a partner.
Got it. John, maybe just to wrap up USIS, non-mortgage, can you help us, you know, quantify commercial and ID and fraud versus the other? So some direction sense of, you know, the faster growing business, like the mixture, basically.
Sure. I mean, ID and fraud is a couple hundred million dollar business, right? And consumers, $100 million-$200 million... Sorry, commercial is a $100 million-$200 million dollar business, and they're growing very nicely, right? And we would expect to see that continue to happen and to become a bigger share of USIS.
Those are like 20%-25% of USIS, way outgrowing the 6%-8%. And then you've got kind of the rest of the businesses inside of that 6%-8% range. And over the long term, we would hope, those businesses obviously, you know, average up, and then when you add the cloud capabilities, you know, if that allows them to move up in that 6%-8% range overall, that's a good thing for Equifax.
And then, is it fair to assume that those are two areas you'll always be on the lookout for? It sounds like-
Hundred percent.
go there, acquisitions?
Yeah, in both areas, you know, we've, we've made, in the commercial data business, two acquisitions, PayNet and Ansonia, in the last couple of years. And then obviously in identity, we bought Kount and Midigator, and, that's, those are... That's differentiated data and, identity and fraud are two of our three, bolt-on M&A priorities. The third is EWS. Strengthening EWS, either in the employer side or in differentiated data around the, talent base. And I think, as you know, we've done a number of, bolt-on acquisitions in the last, three years-
Yeah
You know, to strengthen the core of Equifax.
Got it. Maybe we can shift gears to mortgage-
Yep.
which has not been a friend, clearly. But, you know, John, maybe if you could just remind us of your latest, your last mortgage guidance, but, you know, it sounds like given since the rates have gone up, volumes seem to have come down. So just some perspective on where we stand today versus that calendar.
Sure. For the full year of 2023, you know, we gave a view that we expected originations to be down about 37%, and that we expected inquiries on the credit file, and originations tends to drive the revenue on The Work Number in Workforce Solutions. Inquiries on the credit file, we said we've done about 31%. And what we know is that in the credit file, obviously, what we see a lot more shopping activity. So people pull credit files earlier, so we tend to see more activity, and that's why we see a difference between the activity on inquiries on the credit file and then what we're talking about in terms of originations. And, yeah, clearly correct.
When we gave guidance back in, in July, you know, our expectation was using run rates we were seeing back then, and as you said, we've seen rates, you know, the 10-year up 30-50 basis points. We've seen mortgage rates pop up in the August and September time frame, so certainly that has an impact on the mortgage market.
Maybe as a reminder, mortgage is roughly 20% of Equifax. We overindex on mortgage, I think, as the group knows, because of our large Workforce Solutions income and employment mortgage business. And when you look at 2022 and 2023, you know, last year, mortgage market decline impacted us by about $500 million. The strength of our non-mortgage businesses, which is 80% of Equifax, still allowed us to grow 5% with that decline in mortgage market. This year, the mortgage market is down again year -over -year, and I think John didn't touch on it, but we're at a place now where the mortgage market is 40%-45% below historic levels, what we would call normal. We use normal, Manav, as you know, is pick a time frame.
We use 2015 to 2019. 2019 is before the refi boom during COVID and is a normal market. And when you look at it. It's never been down that low. The market's never been down that low. We've never seen purchase volumes decline to this level. There's a combination of a lack of housing stock. You've got consumers sitting on 3% and 4% mortgages that aren't upgrading. You've got consumers that are moving a lot. I think there's 7 or 8 million households relocate every year. A lot of them are renting while they wait to see where rates are going. When we look forward to 2024, 2025, 2026, you know, that 45% below normal, we would expect that to move towards the mean, towards normal over time.
Now, whether that's gonna be next year or the year after or the year after that, we would expect to see some improvement in mortgage. Then the last point is when you look to 2024, we're not giving guidance yet, but when you think about a market, pick what you think the market's gonna be. Let's say it's flat versus 2023, you know, in a mortgage market. As a reminder, and I think you know this, both of our businesses, USIS and EWS, meaningfully outperform the underlying mortgage market, whether it's going up, down, or sideways. Meaning in a flat market, we'll still increase price, we'll still drive penetration, we'll still roll out new products, and then in EWS's case, we'll add records that drive revenue growth.
In EWS's case, they have a long history of outgrowing the underlying mortgage market by 20 points. So in a flat market, that outperformance will be there, and then does the market improve over time? We would think, yes. The question is over what time frame? In USIS, as more of a single-digit outperformance, but then, when there's meaningful price increases from our partner, FICO, those roll through in USIS and become an outperformance of the underlying mortgage market. So I think an important element to understand, first, is that, you know, the dynamics of our mortgage business, and then second is the power of the 80% of Equifax that's non-mortgage, that's been way outperforming and allowed us to still deliver, you know, kind of mid-single-digit numbers.
What I would characterize, over the last couple of years has been a mortgage recession. You know, obviously, and what's happened to the mortgage market.
Got it. Yeah, I was gonna ask you about that, so I'm glad you brought that up. Maybe just one more on mortgage in the... For the second half of the year, it sounds like there's a lot of your non-mortgage businesses that have a lot of good momentum, and they've had it for the last couple of years too, but unfortunately, mortgage has just been a bigger drag. Is there enough of the non-mortgage momentum to, you know, kind of, offset the mortgage weakness that we might see?
In the second half?
Yeah.
We're not here today to change guidance-
Okay.
—or give guidance on second half, but, you know, broadly, when you look at our non-mortgage businesses, they're quite robust. You know, we're very, very positive about those, whether it's in USIS. International, as you know, over the long term, is a 7%-9% grower against our 8%-12%. They've been outperforming that 7%-9%, you know, for a couple of years.
Yeah.
We feel good about that, and the addition of Boa Vista in the second half will be a positive. We'll probably touch on that later, and then I know we're gonna get to workforce, which is a 13%-15% grower, and, you know, we've been very clear that we expect their non-mortgage growth in the second half and-
Right
Over the long term, you know, to be in that double-digit range.
Got it. Yeah, so let's just move to workforce then. I mean, we addressed the mortgage piece. For the non-mortgage piece, you know, obviously, a lot of questions around the implied assumptions on the government side.
Yep.
So maybe, John, if you could just set the stage on, you know, how much non-mortgage is, and within that, it just help size government, talent, and the other key mixes there.
So non-mortgage is on the order of 70% of Workforce Solutions, and it's grown substantially over the past year, right? So you've seen it's now much larger, obviously, than the mortgage business.
So actually, if you spool back, John, on that 70% non-mortgage at EWS today, if you go back five years ago, it was the opposite.
Absolutely.
You know, so it's really... The non-mortgage has been growing so quickly. Obviously, the mortgage market is taking that down.
Yeah. Absolutely. And within verification services, there's really two big segments, right, of non-mortgage. One is government, and it's over $500 million run rate business, and it's growing very, very nicely. You've seen it grow double digits, you know, 20%-30%, over the last several quarters. Last year, obviously much, much more than that. And then our talent business, which Mark already touched on briefly earlier, is the second largest business in EWS, and it's been growing very, very nicely. Obviously, until early this year, we saw a decline in the second quarter, obviously, as hiring declined substantially, and we covered why.
As we tend to skew toward the higher end of the hiring market, the higher salary end of the hiring market, that's been more heavily impacted in terms of hiring, and we saw some weakness in the second quarter. So, but the biggest parts of that business are really not even financial services related. It's related to government, it's related to the hiring markets, and even in employer services. The biggest pieces of that business, obviously, are unemployment insurance claims processing and processing employer retention credits. The biggest grower, again, is our boarding business, which again, is linked very closely to hiring.
Got it. So let's touch on the government business first then. So, you know, good growth in the first half of the year, but it does imply a pretty big step-up, I think, even in the fourth quarter. So what are some of the drivers there? I know we talked about this last week at the Credit Bureau Day, but some of the full point there, so if you could just help us-
I think, John, you can jump in. I think it starts with a big TAM. Just as a reminder, what we're doing in government, government for us is at the federal, state and local government, we're using our income and employment data to verify eligibility for social services. And you know, think about food stamps, rent support, childcare support. All of those social services are needs-based or income-based, and if you make less, you get more services. So that's, that's what happens. And as a reminder, that's in all 50 states. That's how it's delivered, the social services. And each agency in each state is a different organization that you have to connect to. So we've been building this business over you know, the last five-plus years. As John said, it's a $500 million run rate business today for us.
Last year was up 50%. This year we expect strong double-digit growth again. That's inside a TAM of about $3.5 billion. And when we think about $3.5 billion, the difference between $500 million and $3.5 billion, that $300 billion that's left, is where it's still being done manually by the agency at the social service level. And manual means a consumer comes in with pay stubs and paperwork to the agency office, use the example of a food stamp office, and delivers those to the counter, and they look at them and see, do they qualify? Generally, they'll send them home to get more documents. Using our data, they go home with the food support.
So that's really the digitization of that $3 billion market or TAM is the opportunity for us. And we have contracts at the federal level with some of the large agencies like SSA and CMS, and then we have contracts actually at the state level with each of the different agencies. So that's really the play, driving the penetration there, driving new products. We take price up every year. Generally, those are longer-term contracts, so you have to wait till the contracts open. Where in the rest of our business, we do one-one price increases. But there's just a lot of runway growth potential for us, you know, to get into government. And, John, if you want to add to that.
Yeah, John, if you could just address just second half, specifically the ramp.
Yeah, if you just look at sequentially, right? Things that tend to impact us and benefit us in the second half, we have a substantial contract with CMS, Centers for Medicare and Medicaid Services, around ACA enrollment. That tends to grow as you move through the second half of the year. That's just a seasonal effect. And then also this year, we have, we have redetermination, right? So after the health crisis, now states have to go through and redetermine the eligibility of people, of the people that are on their benefit programs, where they get federal subsidies. And-
Let me just remind, that there was a pause on the redeterminations during COVID. There's a requirement where there's a redetermination every year, so there's, like, two checks that happen. First, there's a check when you join the program and will be ... or most of it's done manually. Then there's the redetermination that's done every year to make sure you're still eligible. That was paused during COVID, and then when the COVID pandemic was lifted a few months ago, there was a requirement to complete a redetermination of everyone during, that's currently in the program the next 12 months. So that's going to span really the second half of this year and into the first half of next year. Then we go back into the annual redeterminations, you know, which is part of the power of the business going forward.
On the redeterminations, you know, states are all doing it at different paces. You know, are they going to do it, you know, in October? Are they going to do it in February? There's some of that taking place, right?
Absolutely. So we feel very good that we're that it's a nice revenue opportunity for us, and we're already starting to see some of it. As Mark said, the timing of that can be difficult to determine because the states will determine when they want to do it. We think we offer the best solution because it's a digital solution that involves a lot of mailing and BPO type services, which are just slower and less accurate.
Got it. And to Mark's point about the contract repricing every couple of years, it sounds like there's a big slug of them maybe coming in the fourth quarter, or?
I would say there's a big slug, but we have great visibility.
Okay.
Like, we know, we know when contracts are up, so when we look forward, you know, we have a lot of confidence we're going to extend the contract. And we, you know, have some visibility around how it's going to be extended. So that, that cadence of contract renewals, and then, as you know, with a $3 billion untapped TAM, where $500 million, it's $3.5 billion market, we have commercial people in the field. Typically, we have them housed or living at the big state capitals. They've got pipelines, meaning they're working on converting customers that are, or agencies that are still manual, that $3 billion going to our digital. And, you know, we have a pipeline that we know how far along those discussions are.
That's stuff we lay in for the third quarter, for into fourth quarter or even into 2024. We have, you know, kind of a cadence of how those are operating and what stage they're in.
Got it.
Similar to redeterminations, we feel very confident they're going to happen. Sometimes the timing can be variable, right? Yeah.
Got it. Okay, fair enough. If you look at quickly, just the talent, I know you guys said you were more exposed to white collar and therefore going to get to lower the numbers there. But, even now, I think it's assumed sequentially, the talent business gets better in the second half of the year. What are some of the non-volume drivers that allow you to do that? Because it doesn't sound like volumes are rebounding anytime soon.
Yeah, it's similar to all of our businesses. You know, you would expect all of our businesses to outperform their underlying markets, the markets going up, down, or sideways. And like mortgage, where we described it, where we get that, call it 20 points of outperformance, same thing happens in talent, where every year we increase price, we add records, so there's higher hit rates. And remember, the power of talent, this is a reminder for the broader group, is, we collect payroll records every couple of weeks from 2.8 million companies in the U.S. We get 50 attributes every pay period from each individual and from the companies. One of those attributes is job title, or what Mark's job title is, a CEO, Manav's is an expert analyst. But we have that job title, and we capture them all.
We've been keeping them for 20 years. So we have a digital resume or digital history on job titles, and as you know, one of the things a background screener does is a part of a background screen is to verify prior employment. It's generally when there's an offer made to employment, the company will then hire a background screener to do a background screen on Mark. They'll do lots of things, a drug test, they'll do incarceration check, which is our data inside of Workforce Solutions. They'll also check where has Mark worked in the past. It varies how deep that is. If you're a blue-collar worker, it might be just last job worked. If you're a white-collar worker, it might be five years of history, it might be last five jobs, and we have that digitally.
If a background screener is not using the data from us, they're doing it manually, meaning they're calling the company, and they're calling Equifax or Barclays and saying, "Hey, does Mark work there? Did he work there?" Companies, the 2.8 million companies that do business with us, when a background screener calls, they say, "Call Equifax." You know, Equifax has the data. We don't do that anymore 'cause we do it for that company, you know, for free. We have 630, I think, million jobs in our dataset. And remember, we have 180 million records every pay period. So we have, you know, a large portion of the working population. So we average about six jobs on the average working American. So that's what we're digitizing in background screening.
So you know, how we grow above market in background screening is the same as mortgage. We take price up every year. Our hit rates grow up as we grow records. We have more jobs. We're driving penetration, so similar to government and even mortgage, if you think about background screening, that's a $400 million run rate business for us, a fairly new business at scale for Equifax. The TAM is, you know, is about almost 10 times that, almost $4 billion. So there's another $3.6 billion of manual effort taking place. So there's still lots of background screeners that aren't using our solution, we're out there converting them from manual to digital, you know, is the other play. And then the last is product.
Rolling out new products for either existing or new customers that have different solutions. Those four are how we outgrow a growing market and how we offset a declining market because those continue to happen.
Got it. And just specific to the second half, right? It's generally around product, and then also we do see generally a seasonal growth as we move through the second half and hiring season, right? And it's very common. So, obviously, new product is something we can drive, will be dependent on seeing what hiring looks like in the second half to see whether that growth occurs.
Got it. Maybe just briefly on competition in Workforce Solutions. I mean, we've talked about it extensively, but in which verticals do you see, you know, the Experian and Truework that everyone talks about? Is it isolated in one, or are they, you know, dipping their feet into all the different verticals, or just some perspective there?
Manav, as you know, we hear about competition really in these kind of meetings, not in the marketplace. I think we've been very clear about that, that we really don't see an impact, you know, from them, where it's not lost on us that Experian has entered the space a couple of years ago. Not lost on us, there's a fintech called Truework that now TU is invested in to participate in the space. But just you got to put a reminder, our business will be $2.5 billion of revenue. I don't know what theirs is, but it's a fraction of that. Both of those companies have some unique records that we don't have, which are valuable. Our scale of our operation, you know, is very, very strong.
You know, to me, a real proof point, you know, in the business is our ability to add records. You know, we grew records in the second quarter, 5 million records, up double digits. And we've had a real cadence, as you know, every quarter of growing records. And we've had the ability to add payroll processors, new, new, new relationships. I think in the second quarter, we added four new relationships. In the last couple of years, we've added, like, 30. So you know, we think we've got a strong market position. We're focused on continuing to grow and expand our business, and, you know, we're investing, you know, huge amounts in technology and capabilities, you know, across all the verticals to, you know, continue to, you know, grow our participation in the market.
Got it.
Our historical data is incredibly-
Very valuable.
Very valuable, and makes it very difficult oftentimes for others who are just entering the business to have that history. And if you think about the talent business, it requires history, right?
Yeah.
The government business, it requires high coverage, or else it's very difficult to get the interest of a government organization, right? So those are markets where without extremely high coverage or, and especially in many cases, historical information, it's very difficult to participate.
How would you respond to, obviously, the competitors are self-serving here, but the idea that, you know, you have all your... Most of your relationships are exclusive, and, you know, at renewal time, maybe they'll be willing to open up to other parties to cover the market or the remaining data that you don't have. Maybe they'll, you know, the other providers will be willing to share. What is, what is the competitive edge with getting all these exclusive guys?
So just as a reminder, 50% of our records we have from direct relationships. So those are from individual companies where we're delivering those regulatory services like I-9, W-2, UC claims, Work Opportunity Tax Credit, et cetera, directly to companies. So that's half our records. You know, that is very difficult to replicate.... Right? So we have a large commercial team, we have a $400 million business. You know, and those relationships, again, we do income and employment verification for those companies for free, so it's tough to compete against. And we have those services, you know, that become very embedded in the company. So we feel very good about those.
And, you know, we've had a very strong track record as long as I've been here, where when one of our partners, 50% of our records come from partner relationships, and it's payroll processors, HR software companies, pension administrators, tax preparation services. So a wide array of different partner relationships that we have. We generally pay a rev share on those. Those when the contracts come up, they extend. And your question is why? It's the track record, it's the scale of our capabilities, you know, the regulatory element. You know, if you're a partner, you don't want to be involved in the regulatory side of this. We do it really well. But I think it's our 20-year history of being in the business.
When I meet in the C-Suite of partners like that, you know, their dialogue is mostly around, "Don't screw up my core business. We wanna make sure you're gonna do this really well." And when we've invested $400 million in our tech stack over the last, you know, three to four years, they have a lot... We have a lot of credibility about, you know, doing it, doing it really well. So those contracts, when they come up, we find them extending. And, as we pointed out, we have the ability to monetize their records over a wide array of verticals, you know, government, talent, financial services, auto, cards, P-loans , and then, of course, in mortgage. And those are different integrations.
You know, we have tens of thousands of direct customer relationships that, you know, again, have been built up over 20 years. It's a part of the scale of the business.
Got it. Fair enough. I want to quickly shift over to margins. You know, I think the fourth quarter margins implied 36%. I was just wondering, embedded in that run rate, John, maybe two parts. One, can you remind us of the, the cost-saving initiatives you took this year and how their run rates and what's left to, you know, further, help next year? And then the second question I'll follow up on is just more, you know, that USIS cloud and the transition.
Sure. So we announced $210 million worth of spending reductions that would impact 2023, and then in 2024, there's an additional $50+ million. So, so it carries into 2024. About, about 60% of that spending reduction benefits cost and about 40% benefits capital. The bulk of the actions get executed during the year. They didn't really start in the first quarter. They ramped in second and third. So obviously, we're seeing almost the full run rate of cost benefit. Not quite, but almost the full run rate of cost benefit as we get into the fourth quarter. And that really does benefit substantially, obviously, our, our, our margins as we look into the fourth quarter.
Also, what we're seeing is we're expecting to see, as you've talked about, some improvements in revenue as we move through the fourth quarter. So as that happens as well, then what the very high variable margins we have, as well as the cost benefit we're getting from these actions, is what allows us to drive the margins higher.
Got it. And then just on the cloud, maybe first, what is the latest timing in terms of which, you know, business line-
Yep.
-country, when do those sunset?
Yeah.
The deal comes to, yeah.
No change from our recent dialogues on that. First off, EWS is substantially complete. I think the real positive is we're seeing the benefits of EWS being in the cloud. So they completed last year, and I'll use one metric on new products. The group knows we have a metric we share with you, and we really use internally called a Vitality Index, which is a percent of our revenue from new products introduced in the last three years. Historically, we were 5%-7% vitality. When we made the move to the cloud and we put our new long-term framework in place in 2021, we moved that to 10% vitality from 5%-7%.
Just, I think, as you know, we were 13% last year, we were 14% in the second quarter, and we expect to be 13% this year. And, you know, while we're not Apple, we think that's a very vibrant company that has more than 10% of its revenue from new products that weren't around three years ago. So that's, you know, the focus. And I think, as you know, new products generally drive higher price points, which means higher revenues, so they drive our top line, and they also drive margins, because that incremental growth is very high incremental margins in a business like ours. You know, our incremental margins are in the 78%, so very attractive margins. So back to cloud, EWS completed last year. Pre-cloud, you know, they were south of the 5%-7%.
They were 3%-4% Vitality. Post-cloud, they're over 20%, so they're almost double, you know, what we're performing. And that's a real proof point for us about all the many benefits of being cloud native. We were convinced it was gonna allow us to innovate more, which would bring more solutions to our customers, which would drive our top and bottom line, but also make us a more important partner to our customers. So EWS complete last year. USIS will be complete, call it, you know, midyear next year, which is a big milestone for them to get complete. That was a more complex cloud migration for us because of the, you know, scale of the data assets that they have. You know, EWS really has TWN and our insights data. USIS has multiple datasets, so they'll be complete next year.
A handful of our international countries will be complete by year-end. Spain, some Latin American countries are complete by year-end. Canada completes, which is similar to North America, in first/second quarter next year. U.K. in the same timeframe, and then Australia will follow. So, you know, in 2024 and 2025, like a year from now, when we're back at your conference, we'll have North America done, and we'll have much of international done, is really a big deal. And you asked earlier about cost savings. You know, as John pointed out, we got the carryover this year of the cloud cost savings from completion that we have, and then, that continues in 2024 and 2025 when we finish it up. And remember, part of the cloud cost saving is we're running different environments today.
You know, in USIS, we had the legacy mainframe infrastructure still operating, plus our new cloud environment as we move customers from legacy up to the mainframe. So we're super energized. I think the other, you know, power for us is that for the last, as you point out, long time, four or five years, we've been running the business and migrating the cloud. Not for the faint of heart, a super hard project, but everyone in the companies had basically two jobs. When we complete the cloud, we really focus on one thing, running the company. So that's gonna be a really, you know, big change. These are not easy exercises. We're convinced it's gonna transform Equifax and give us a competitive advantage, you know, for five years, ten years, you know, versus our competitors, because they're, they're doing a different version.
You know, they're kind of going slower. They didn't have the window to go fully cloud native that we had. We think that's a big deal for Equifax, you know, as we move forward to 2024, 2025.
Capital comes down at the same time, right?
Yeah.
So capital comes down to about 7% of revenue and could potentially push below that, and then also significantly increased percentage of our capital moves to product as opposed to transformation or infrastructure.
Yeah.
So it really does help accelerate our NPI process, because we can shift funding there, and as Mark said, also shift the focus of people there as opposed to focusing on transformation.
As John pointed out, as CapEx comes down and our free cash flow and our EBITDA expands, you know, as we go forward to 2024, 2025, 2026, our free cash flow expands. When you get out to a year from now, 2025 and 2026, we'll have significant excess free cash flow from that lower CapEx and higher margins that we intend to return to shareholders.
Got it. All right. I think we're just almost out of time, so we'll stop it here. But thank you, Mark and John, for being here, and thanks everyone else as well.
Thanks, Manav.
Thanks, Manav.