Flywire Corporation (FLYW) Earnings Call Transcript & Summary

Thanks for joining us. 9:15, we're going to get started here. I'm Michael Infante, I'm an analyst covering fintech here at Morgan Stanley. Very pleased to be joined by Cosmin Pitigoi, Flywire's CFO. Before we get started here, I just have a quick disclosure to read. For important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. So with that out of the way, Cosmin, thanks for joining us.

Awesome. Thanks for having me.

So maybe before we get into the details, just stepping back to 2025, obviously, a lot of external noise around international student visa trends. You guys still ended up delivering 17% organic ROS growth and an average of about 6% estimate upside through the course of 2025. Maybe for those in the audience that are a little bit newer to the story, can you just sort of unpack what your initial assumptions were for revenue growth in some of your core education markets in 2025 relative to the results that you ended up delivering?

Yes. And I think to some extent, maybe we need to reintroduce ourselves a little bit because I think if you look over the last few years, we've just become -- and I've heard it from a lot of folks that we've been so much more diversified than if you think even 3 or 4 years ago, everyone thought of us as sort of a cross-border mostly education sort of business. And now we're diversified beyond just not just the cross-border business into other verticals that have become a bigger portion. We've also seen a lot of growth on the domestic side, increasingly enterprise sized clients and increasingly more -- just more and more software than in the past. So we've certainly evolved from that and increasingly just talked a lot more about free cash flow and margins continue to do quite well. So that's sort of the backdrop of kind of how we've evolved. Now if you think, yes, to your point, so early 2025, certainly a lot of uncertainty in the environment. And so what we did was we took a very prudent approach given that to the guide. And listen, at the time, what I said is I'm going to take sort of 3 principles. One is prudent; two is data dependent and very transparent. And so what we did is, yes, we assumed both Canada and Australia will be down around 30%. Now what we saw there is Canada was a little bit better than that, even though actually the visas were down more than that, we actually performed better than that. And then at the other extreme, Australia, where, again, looking at early trends in 2025, it looked like maybe it could have gone the same as Australia -- the same as Canada, it did not. So actually, Australia was a growth market for us in 2025, and their visas actually are relatively flat. And so those are some of the drivers. But look, at the same time, what we also did is we took -- we assumed U.S. was going to be down. It didn't ended up being actually better and U.K. continued to outperform. At the same time, though, we really emphasized a lot of the sort of fundamental aspects of the business. So we did the restructuring. We announced that we're looking at sort of everything in the business, and we executed on a lot of those margin and cash driving components, which, again, these are the things that focus on what we can control to show the value in the business. And that's, I think, as you exit the year, we're not just more diversified, more of a margin-driving business, but also a much stronger kind of foundationally business exiting the year coming out of that.

That's helpful context. As we think about 2026 and the outlook that you just gave to investors, market by market, you expect U.S. visas for first year to be down about 30%, Canada to be down around 10% and U.K. and Australia to be flat, respectively. I know we're early in the year, but maybe you can help sort of investors understand how you sort of formulate some of those assumptions. What are you seeing on the ground in your own deposit data? What are the conversations with agents look like, et cetera?

Yes. So look, taking the same approach, again, sort of very prudent data dependent and transparent. So when we look at some of the assumptions around the U.S. being down 30%, right, we look at -- we talk to agents. Of course, we have our own agent networks. We talk to them. We look at common app data, which is available out there. So what we're hearing on the U.S., for example, is -- and you've seen this in the F1 data over the last couple of years, some pressure in the India channel. So sub students from India choosing to perhaps go to other locations. So we're seeing diversification, which again works kind of in our favor, too, because we have the footprint globally. So we're seeing that a bit play out. It was already playing out in Canada. We saw that. It's -- and it's been playing out in the U.S. So -- and we look at common app data. It's a good indicator. Again, a lot of these things are indicators until you kind of see the actual seasonal peak play out. It's kind of tough to call how these things play out, and that's kind of the data-dependent piece of it. But look, at the end of the day, we're continuing to take a very prudent approach around it. And again, in Canada, we're assuming some small pressure there, but you've seen in all of these cases, we've taken a pretty prudent approach around the assumption, and then we'll see how things play out, and we'll update everyone. But along the way, we are seeing students apply to more. So I guess the summary is students are applying to more destinations around the edges. Obviously, the big -- there's still -- there's a lot of demand for still international education in general because you're always going to have an imbalance of talent in different countries. And so I think -- and I still believe as -- including myself as an immigrant a little bit, but believe in the strength you build in your ability to attract the best talent to the country that you want to kind of -- especially as we're AI conference, tech talent and talent in general, I think, will be still a demand for it, but I think the diversification of those is kind of what we're seeing, especially from the India channel.

How about from the actual spread between your own visa assumptions and the revenue growth that you end up delivering, right? So if we break it down market by market, the spread between the U.S. visa degradation and your assumptions for U.S. revenue growth, it's in the low 30s percent range, right? If we take the Canada market, it's closer to 20%, U.K. similar in Australia, a little bit lower. So can you just sort of unpack why the spread between visas and revenue in the U.S. is so large and maybe why it's so narrow in Australia?

Yes. So let's start with that. So if I start with the U.S., that's our longest and most mature market. And so we've built many years of cohorts of students over time. And so the things that play into that are some of that retention of some of those cohorts as -- as you would realize, obviously, the visa component mostly impacts your first year kind of cohort. So the other cohorts and the retention of those offset some of that pressure. So when we talk about last year, assuming visas were down 20%, again, we're still waiting for the data, but just looking at our own data on first payers, so it's down in the sort of high teens. Offsetting that, we do see this retention aspect where about half of our revenue in cross-border U.S. comes from existing kind of cohorts and the other sort of less than half is from new. And that enables us to then have improved or lower sort of impact from that. And then, of course, you have the tuition aspect where undergrad is still pretty demand, so higher tuition kind of opportunities. But also, we have other levers, right? So we continue gaining share. So we announced continue to win in cross-border. And then we have the domestic aspect in the U.S. where -- especially, I think we've sort of given these metrics, but specifically for 2025, if you look at the U.S. business, the domestic portion is in kind of low teens share of revenue -- sorry, the cross-border business is low teens share of revenue, while the domestic is sort of mid-single digits. So that domestic portion is growing much faster and also, I'd say. So again, continuing to win enterprise SFS clients. So all of those things play out into that spread for the U.S. In Australia, it's a bit different. They have not seen as much even though we thought they would, but they haven't seen as much pressure on the Visa side. But we are -- we have larger clients there in particular. So -- and we also have an insurance business, if you recall, the Cohort Go acquisition. So that is impacted by sort of volume of students -- so all of that kind of plays into Australia having kind of a narrower spread. Although if you think of 2025, if you assume visas are roughly flat in Australia, and we grew in the low teens. But again, Australia, we did -- the market did better than we thought, and we also did -- we had kind of a higher spread. But again, as we get into this year, we want to be prudent. Look, U.K. and then U.K. and sort of Canada is kind of a different aspect where U.K., we've continued to outperform significantly. The Visa market because we've got product is great, lots of interest, very different competitive dynamic. And then in Canada has obviously just been very volatile. So we've continued to win share there. And we -- and now as they've come off of this volatile kind of time, we expect, as you can -- as you saw, with Visas down 10% to continue having that spread that kind of works in our favor. But that's where, again, it's sort of structural and share gains and more so than anything driving this.

Before we pivot to SFS, I just wanted to hit on the U.K. market in particular. One of the concerns that we tend to hear from investors about the U.K. specifically is sort of just looking at the major schools in terms of the composition of international students and investors coming to the conclusion that you already have a fair amount of those universities. So what would you sort of say to investors that are concerned about the client penetration levels within the U.K. and how you're thinking about incremental revenue growth in that market?

Yes. I mean, listen, if you look, we've heard this sort of question for several years. And to some extent, we did disclose the U.K. this year for the first time to show that it is not only our largest education market, but also growing at 25%, which is obviously ahead of company average. So still a lot of growth strength there, given the product penetration that we have there with the domestic business, I think -- and now with the cross-border and domestic together with SFS. And so I think the right way to think about it is if you just look at number of clients or schools, what you're maybe missing is our penetration within those schools. And so the stat that we -- and so for us, there's the opportunity to sort of call it, if you think of it a share of the student financial flows at a particular school is more important than you could have all the schools, but if you don't have the full share within those schools, then that's less relevant. So for us, the one metric to keep in mind is last year, I think we said about a dozen or so schools and now there's a bit more. We have about 90% of their student kind of financial flows. And so if you think of that, obviously, a dozen is obviously a pretty small number generally related to the entire market over there. So for us, that opportunity in the U.K. remains to continue gaining more share within those schools because we have those integrations that are very unique that no one else tends to have. So that's the way to think about it, and we'll continue to unpack that, but we feel the U.K. should continue to grow at pretty healthy rates. For this year, we said U.K. plus EMEA will grow at or above the company average.

Helpful. On the SFS product, what percentage of your education clients are using that product today? And how should we be thinking about SFS penetration as a percentage of education revenue and where you might expect that to go over the near to medium term?

Yes. In the U.S., we've given the stat where we have about 1,000 institutions on our sort of legacy cross-border. And by the way, we keep adding that. And then of those, we have about a 10% attach rate. So about 110 signed 13 SFS clients last year in the U.S., which is a record number for us, and we continue to see strong pipeline demand there as the perception of us has shifted from -- it used to be, again, it's not just, I think, the perception for, to some extent, shareholders and investors and others, but also, I think, for our clients is shifting from, oh, you're not just the cross-border provider, you can also do domestic. So those conferences, and we had, again, our conference this week in the U.K. We have a similar one in the U.S. So we see that capture. But 10% adoption is still early. And in the U.K. is even sort of lower in the U.K., we started with. We have now 6 live on SFS with 4 kind of design partners. So where we kind of test the product and launch and the demand there is quite great. And again, U.S., we differentiate on a number of product and innovation capabilities. And so we continue to see a long runway there now that especially that kind of -- that mindset has shifted around kind of how people view us. But continue to expect, like I said, in the U.S., even this year, if you're saying low single-digit growth despite a 30% visa, you have to assume that the domestic business is assumed to be growing pretty healthy. And again, that's us, again, remaining quite prudent around it, but the team is quite excited around the opportunity on SFS. It's a longer cycle, but we like it because then it's -- those clients stick with you for a very long time.

On the scaling cadence, how important is ERP coverage there to sort of scaling SFS? And what do you have to do internally from an integration milestone perspective to make sure that those integrations are in place and you can continue to get the attach that you want?

Yes. I mean the integration is an incredibly critical differentiator and that last mile of any payment that we always talk about is in many -- and for us, there's a last mile in both directions. There's the last mile as far as the university receiving that money and having to book it into a system of record and that system of record many times, again, we gave this picture of the U.K., for example, we have 4 or 5 different types of integrations and building those, whether it was Tribal, Unit4 (Aggresso), and now increasingly, you've heard us talk about Oracle Fusion. All those are super critical, and they're difficult to build, and you have to actually work with those partners. So that creates a very unique differentiator for us. And then, by the way, on the customer side, providing the banking relationships on that side and being able to connect, again, that last mile on both sides is a big differentiator, but the ERP piece is very critical. Differentiator huge in the U.K. In the U.S., other -- our competitors have that also, but that's where kind of our product and innovation and sort of that's kind of what's been driving our differentiating factor there, too.

As students establish local banking relationships in years 2 to 4, you obviously have really healthy net revenue retention metrics even with the macro headwinds that you saw, still delivering 110% plus NRR. How do you think about that payer retention piece and just the dynamic of students establishing those local banking relationships and what you internally can do to drive higher share of wallet?

Yes. So NRR increasingly now as we diversify is obviously, there's a number of components driving it. But within education, we still see healthy retention and NRR because, one, you have on the enterprise side, obviously, you have very low churn, which plays into that. But yes, there's a few things that drive sort of student retention for us, and these are levers that we obviously push quite a lot to make sure that we do create these long-term relationships with the students. So first, SFS adoption. So that is as you go from being just a cross-border provider to then processing kind of all the student flows, I think you see and we talked about it, and there's a few slides you can look at as far as how that plays out. But imagine it's sort of a 3 to 5x revenue multiplier and sort of a 2 to 3x gross profit multiplier as you shift. But we do see just more retention even in the cross-border students from first to second year because of that -- because we now sort of process their payments. Then the second thing is managing sort of presentment at checkout. Obviously, we can always optimize that and provide a better experience for the student. And that's another way to obviously help manage that sort of retention. And then third, which again, is kind of unique to us is partnerships where our partnerships with the banks upfront. So that was kind of my point around the last mile on that side, where we have partnerships with the biggest banks in India, with banks in China and other places. So creating that sort of level of comfort with the student and knowing the name and feeling sort of safe with obviously your life savings in many cases that you're transferring creates that sort of, again, retention in future years for us.

That's helpful. Maybe pivoting to more thematic topics with both AI and stablecoins. As you sort of think about the opportunities and risk to Flywire from an AI perspective, how should we be thinking about things like your routing economics, some of the reconciliation productivity the efficiencies that you're seeing with the sales organization and customer support. Maybe just sort of like rank order some of the buckets where you see AI being a tailwind to the business and/or areas that you think sort of reinforce the stickiness of the platform?

Yes. I mean, as I said, I'm a data platform and machine learning sort of guy for over 20 years. So these are exciting times for me in terms of what you can do. I've always tried to automate myself if I can. And in every situation, and these are -- now you have the tools and the ability to move even faster. And look, I think of it as AI as sort of there's the fluency and then there's the accuracy aspect of it. So wherever there's a number of documents that need to be summarized and synthesized, those areas are going to be very quickly automated. So for us, customer service is obviously the easiest first place to go. And so for us now, I think it's something like almost half of our calls are through kind of some sort of automation. Routing economics, although obviously, AI can accelerate this, that's been something that we've -- as being in payments for 20 years, that's been an area of always an optimization through a number of different models. You can think of any sales also, we've now implemented a number of tools for the sales team. So if you're managing multiple relationships and documents and obviously, you can automate. So those are the fluency level kind of areas are one aspect. Then there's coding itself. Obviously, our engineers can be used, I would say, -- majority of our teams are basically use AI-enabled almost for everything, for every coding kind of activity that is kind of AI-enabled. And so they already are accelerating everything they're doing through that. So that becomes kind of an opportunity to continue optimizing the R&D function, if you will. And then on the G&A, and I think the last bucket where a lot of my passion and history is around data architecture and data governance, where rebuilding and improving your data architecture underneath kind of your systems of record and rebuilding the systems of record and consolidating that, then you can actually build agents on top that can communicate with all of that data at scale. So if right now, your HR data is separate from your finance data is separate from your treasury and so forth and/or marketing and risk and other areas, you missed the opportunity to really see end-to-end. But part of that is, one is data architecture, two is data governance, that you set the right definition. So when you send an AI agent to say, "Hey, give me the revenue for this market, this product, tell me the risk related to it and did we run a marketing campaign in that region?" Like all of these sort of definitions are key. And so I see that as improving the G&A function then even faster, but it requires some investment. And that's one of the things that I've done in my career in my past, but we're doing here, too. So you've heard me talk a lot about that investment in terms of the digital transformation architecture.

That's helpful. On the stablecoin topic, I was actually somewhat surprised to hear Mike talk about the demand that you're already seeing for stablecoin-based payments. How are you thinking about that more as a customer-facing payment method versus more of an internal segment rail where you have an opportunity to sort of reduce some of the payments and settlement-oriented costs, where you think we'll see the impact show up first?

Yes. Look, we're -- while we haven't heard a lot of our clients come to us and say, I'm going to -- I want to pay my tuition or pay for this very large sort of vacation that I've saved up my sort of life savings for with stablecoin yet. We have a partnership with BVNK and are already testing that. But -- and the way we think of this is yet another rail that we can enable for our clients. Look, this will be actually incremental for us because if you think of countries where it's either high inflation or volatility, this could be an actual incremental rail for us that is helpful. So we'll be ready for it from an economic standpoint. It's not very different than sort of our lower cost. But I would say the thing on this is kind of the main takeaway is that we -- the rail is not the value that we provide. It's the orchestration and the integration with the client that really matters, and it's that end-to-end workflow. And that's kind of going back to the -- whether it's the AI conversation on the software side or stablecoin it's the end-to-end orchestration and integration, and we're solving that workflow kind of the workflow point and in a regulated industry that really matters. So the compliance, the regulatory, the FX, the integration are all the things that our clients pay for, and that's the value we bring more so than just the rail. But we'll be ready. We -- our Chief Payments Officer has been on this, and we're excited to continue unveiling kind of as we innovate in this area.

Helpful. On that sort of disintermediation topic, any updates on just the overall competitive backdrop? We obviously, at this conference have widely heard about the pace at which new businesses can get formed and sort of try and attack some of these industries with more AI native solutions. But how do you sort of think about the stickiness of Flywire. You obviously mentioned the breadth of the payments network and the infrastructure that you have developed. But how should we be thinking about that structurally?

Yes. Look, I mean, I think there's a number of different scenarios that we've all played out. And we -- to me, it's interesting that we've had better software for decades and always had that argument to say, we've got better software than the incumbent. And so I'm excited to hear that the better software should win faster. So for us, that actually works in our favor. And look, we've looked at it. You can look at it from a checkout, from a stablecoin, from the banks and the ERPs, who's going to play in that space and whether it's -- look, it's not just checkout. Checkout is one step in the process, as I said, is the -- being able to -- so you can by code the checkout maybe experience, but then you -- once you receive that money, if you're an educational institution and great, now you've got $10,526.51 from somebody -- some name, now you got to go reconcile that into your system of record to the penny. That is a differentiating. That's one of the many steps in the process that we solve for, which, again, would not necessarily be solved by just the checkout experience. The same with -- again, with stablecoin, we sort of talked through the rail kind of example. And then look, on banks and/or ERPs, banks are not necessarily going to build software and the vertical integration and domain expertise because we are in very unique domains that require very specialized knowledge to even interact with, whether it's a school or a travel operator. And so again, and we're not standing still here either. And then on the ERP side, most of those folks don't really want to deal with the payments and regulatory issues and licenses and all that stuff that comes with it. So -- and we -- obviously, we play in those, and we've been playing in those areas. And obviously, we're not standing still. We keep moving quite fast in terms of our own innovation. And we are in very unique verticals that play to our strength. And the last thing I would say is we are focused on that accounts receivable piece, which is also a little bit of a differentiator. We can do the payable side, too, but the AR side of it is also kind of a unique thing. So I know we're hard to compare to anybody else, and that's the way we kind of like it. But we feel like we -- this is an environment that actually works in our favor as kind of being the disruptors.

Helpful. On the gross margin topic, you obviously have several puts and takes this year. You also have the structural dynamic where things like B2B and travel are growing quite quickly and are, therefore, a gross margin headwind, but still dollar accretive. How should investors be thinking about the ramp that we're seeing this year in terms of domestic payments processing with the client ramp that you mentioned? What's specifically driving that drag? And when should we be thinking about that outsized headwind sort of lapping that dynamic?

Yes. And one of the things going into this year that we wanted to do is also remind people that, look, what really matters, yes, we look at revenue and gross margin, but gross profit dollar growth is actually one of the key metrics, and you've heard Mike talk about is one of the areas and metrics that we're focused internally and balancing that gross profit dollar with every dollar of revenue is key. So I would say that is not so we even sort of directionally guided the mid-teens kind of gross profit dollars. So -- but sort of the dynamics on gross margin are sort of threefold. One is kind of mix, which we've always talked about, so it's business mix. Second is FX. And then third is this temporary kind of payment processing ramp. So if I unpack those briefly, those 3 in Q4 last year, you saw about a 600 bps decline. And the breakdown is about 1/3 across each. So there's about 1/3 of it was kind of our normal mix of business, about 1/3 was this FX on settlement, which we'll talk briefly about. And then another 1/3 was this payment ramp from Cleveland Clinic, in particular, and then the B2B invoiced cross-sell that we've started. So the way to think about the 3 components, the mix is mainly driven by domestic and EDU growing faster, so SFS and others, still positive gross profit growth, as you saw. And then whether it's travel or B2B, those growing faster than the company average with a lower gross margin. So it's not sort of pricing pressure or anything like that. It's sort of spreads are quite stable, as you saw in our supplement. And so it is just -- that mix is kind of a function of just business mix. So overall, that -- so then that's the part that will continue to be sort of 100 to 200 bps into next year. Then FX is actually 2 pieces to FX. One, gross profit is at spot. So we're not -- but it has very similar dynamics as kind of the revenue side somewhat. And then this FX on settlement is -- and we've talked about this briefly before, is the timing between when we pay a client and we get paid. And it's a few days, and we hedge that on gross profit on the cost of sales line, but then the hedge shows up in OpEx. So actually, on EBITDA, those are pretty well hedged. So what happened there is -- then in Q4, what you saw is a bit of a headwind for about 2 points because the prior year kind of versus year-over-year impact was impacting that. And what you'll see this year is first half actually will reverse and it will be actually a tailwind because last year was kind of a negative in FX on settlement. So first year will be a little bit positive, a couple of points or so. And then the second half will be sort of offsetting that. So for the full year, that kind of neutralizes. And then last thing, again, payment processing, again, we kind of unpacked it in the guidance to the point of being transparent. So again, more of a first half by the time we get to Q4 and exit Q4 should kind of normalize back into that 100 to 200 bps kind of range.

Helpful. On free cash flow generation, obviously, I know it's a priority for you in terms of ramping that. Can you just talk through some of the nuances with the 2025 free cash generation? I know there were some internal initiatives to sort of prepay expenses to achieve more favorable rates on a go-forward basis. How should we be thinking about what you're expecting in terms of the ramp of free cash flow generation as that prepayment dynamic normalizes?

Yes. I'll say 2025, the lower conversion versus -- we guided 70% to 75% for this year, which implies, yes, I realize a pretty big jump versus '25. But '25 was depressed by a number of things. One was the restructuring. So that was a big chunk of it. Second, we did have some M&A costs and related costs that were sort of onetime in nature. And then third, yes, there were some prepayments where we can lock in a contract for longer by prepaying and getting a better rate. So we did take some of those there. So when you actually unpack all of that, you get back to a similar range of already kind of a sort of normalized basis, even last year was kind of above 70%. So then getting to 70% to 75% this year doesn't become quite as big of a stretch. And in addition then, we have an opportunity to -- as we move from just revenue and EBITDA to really look at gross profit, free cash flow, you heard me talk more about the importance of gross dilution and net dilution and now free cash flow, all these metrics that are kind of foundational are something that we're going to look at. And for free cash flow, working capital, obviously, there's lots of opportunities there to also tighten up the working capital levers to continue improving. But 70% to 75% for the year is kind of the right range for now.

And lastly, from a capital allocation perspective, obviously, with the expected ramp in free cash flow, how are you sort of thinking about the decision framework about reinvesting in the product, things like ramping SFS penetration, et cetera, relative to continuing that sort of historical acquisition cadence that you historically have been on as well as just the buyback itself. How are you sort of thinking about the puts and takes across those 3?

Yes. I mean I think we're still in that same order. So organic investment first. So whether it's the domestic penetration in SFS or travel is a key area, obviously, a combination of Sertifi and legacy for travel luxury for us. And then third is kind of this more foundational investment in kind of AI and data architecture. All those are key, and that's number one. Two, buyback, as Mike sort of said on the earnings call, obviously, the most dislocated thing we see right now is kind of our own stock. So it will remain kind of -- the buyback will be a big focus. And then third, look, we'll continue to be looking at the M&A market. But obviously, we have the Sertifi acquisition, to some extent, the Invoiced acquisition that we are focused on from an integration and getting the synergies there. So that will be a big focus, but certainly always looking in the market to see if anything interesting comes up. But yes, the first 2 are going to be our main focus.

Awesome. Well, Cosmin, I think we're going to wrap it there. Thanks very much for joining us. Appreciate it.

Thank you.