Good morning, everyone, and welcome to day three of the 2025 Stevens Annual Investment Conference here in Nashville. We're excited to have eXp World Holdings take our eXpI with us today. Joining us are Glenn Sanford, Chairman and CEO; Leo Pareja, CEO of eXp Realty; Jesse Hill, CFO; Wendy Forsythe, CMO; and Denise Garcia, Head of IR. For those new to the name, eXp is a cloud-based real estate brokerage, and over the past 15 years, it's evolved into a global platform company that combines real estate brokerage, technology, and community at scale. We'll start with a brief company overview, moving to strategy and platform evolution, then discuss agent growth, international expansion, financial performance, and finally, the broader housing and competitive backdrop. Glenn, let's start with the basics. For those less familiar with eXp, can you give us a quick overview of the company, your core business model, scale, and what differentiates you from traditional brokerage networks?
Yeah, thanks, Oscar. Yeah, so eXp is a company I founded a little over 16 years ago, and it was really leveraging technology in a really bad housing market, recognizing that the future of real estate brokerage would eventually be without the sort of bricks-and-mortar component, just sort of recognized where high-speed internet would take everything back in 2008, 2009. It hadn't been built yet, and we said, "Okay, why don't we build it?" For a couple of reasons. One, we couldn't afford offices, but two, we just recognized that that was the future. Starting with 30, 40 agents back then, which was my real estate team, we've now grown to somewhere close to 83,000-84,000 agents in 28 countries, and we have fundamentally changed what it means to be a real estate brokerage.
We see real estate agents as being the individuals who generate the clients for the brokers. They need a platform that they can trust, that's innovative, that brings them sort of the best tools and services, but also the ability to create community, create referrals, build additional sort of ancillary relationship-based businesses on top of this platform. I think we get sort of put into the bucket of a real estate broker, but we're really a technology platform that is disruptive, similar to how Amazon sort of disrupted retail, Netflix disrupted video, Uber's disrupted transportation and taxi. We've really disrupted the legacy bricks-and-mortar-based real estate brokerage play and got rid of really the bricks-and-mortar tax on real estate professionals.
Great. To your point of being a platform, what does that mean operationally today, and how do initiatives like Connect Hub, Success, Frame VR.io, and others fit into that vision?
Yeah, so the one thing, if you think about being entirely cloud-based, the big thing is how do you build collaboration and community? You've got even companies today that are trying to bring people back to the office because historically, that's where community and collaboration took place. We recognized in an early stage that we needed platforms for community. Frame VR, it's a metaverse platform. It's spatial computing. It's really where a lot of technologies—Apple just released their Apple Vision Pro. You've got Oculus continuing to invest, Steam investing in these virtual environments. The virtual environments create a way to create community. Community is a big piece of the whole infrastructure. It's the way that we collaborate. We've now started to even build our own internal collaboration tools using AI and even localizing that to, I think, about 13 different languages. We have done it all in-house with some of the AI technologies, which, again, lends itself to low-cost, high-impact for our agents and then just sort of the overall global footprint.
Great. Shifting to the agent side, maybe for you, Leo, roughly 40% of new joiners were team-based in the third quarter. How does that structural shift impact revenue share dynamics, margin mix, and just productivity overall?
Yeah, so if you go back to the origins of the company, Glenn Sanford was a mega team leader, right? He started as an agent in Keller Williams and then had a multi-market strategy with Buyer Tours. The DNA of our company started with mega teams, and our platform lends itself specifically well to teams. The friction that exists in the old legacy competitors in the franchise systems creates a massive amount of friction for teams that are expanding quickly. Think about if you were to be associated with a RE/MAX or KW franchise, and you want to go to the neighboring county, you all of a sudden have to negotiate new economics. There are problems with the geographic boundaries. Our teams are multi-market, multi-state. Some can even cross country lines, right?
We have teams that operate in Puerto Rico and Florida, which, again, is a U.S. territory, but for all intents and purposes, it behaves like another country. Because we're a centralized, independent, singular platform, it allows specifically for multi-market expansion. If you were to look from 2000 through today, the size and the units done by teams continue to go up. Even if you were to take a year like 2010 that had the same number of transactions and you were to look at the leader on the RealTrends list or at the top brokerages, that's grown by an order of magnitude. We have focused there for a couple of reasons. One is they're substantially stickier. Teams or individuals that do over eight transactions churn substantially lower than the rest of the cohort.
Teams over 21 transactions are super sticky at about 2% annualized, so like 0.13 per month. The unit economics on teams is still pretty good because they're so productive. They fully cap, and then we have post-transaction fees with the model. They also attract the aspirational agents. In a down market like this one, it's very much a land and expand and take market share, even though ourselves and everyone of our peer group have experienced downward pressure. This is the time to expand market share, both in listings and productive agents. As hopefully there's a recovery back to normalcy and defined as somewhere between 5 million and 6 million transactions, we will get the buoy effect of having the lower production agents in the bell curve join us again. We're hyper-focused on making sure that we're serving their needs. The higher their volume, the harder it is for them to make the economics work at a traditional company that's uncapped.
You've touched on stickiness. Obviously, with the challenging market that we've had over the last few years, the entire industry is facing a lot of attrition, yet eXp was able to improve attrition 18% year- over- year. Which are the initiatives that are helping?
Yeah, so that's the fun part about business. That's the two-year overnight success, if you will. There are a lot of initiatives that we started a couple of years ago. I would say that as we saw competitors copy our model. To Glenn's point, we are Netflix. I would point out that Netflix disrupted how content's created. There is no new startup MGM buying land, building a studio, and then selling monthly subscriptions via cable company. You have Hulu, you have all the other competitors. They had to invest substantially more money to not capture the same number of eyeballs as Netflix. By being so, we've created a margin advantage, especially in our direct competitors. That's sizable, 40-50% more on the top of gross margin if you look at Q3 reports from all of our competitors.
We have been able to take that money and reinvest back. For example, we stood up eXp University about two years ago with the specific goal of giving them content that increases per-person productivity. A fully aligned thesis in where if I do not charge for coaching and services like other franchise competitors because they do not have the same revenue profile, we have the ability to give it away, not because we are purely altruistic, it is because it actually drives per-person productivity. For a team, the way we look at the value stack is what can we take off of your P&L, of your OpEx, where we can add value and become very sticky.
Putting all of that together, you have people joining as part of teams, you have improved attrition. Glenn mentioned earlier that community is a big part of this business. In a virtual brokerage model, what are the most effective ways to maintain that community, collaboration, and loyalty at scale, like you mentioned?
Yeah, so I think one of the misnomers is we are completely office-less. The top teams tend to have physical locations in their regions and open them up and use them as a form of culture. We programmatically have lots of events. In 2024, we had eXpCon Canada, eXpCon Barcelona, followed by eXpCon Miami. In addition to that, we had dozens of rallies throughout the country. For 2026, I think we were up to 36. We, at the corporate level, have hire and masterminds every 90 days with our top performers and top attractors. At the local level, we'll give them playbooks on how to host events that are production-driven.
We're just coming off of an internal meeting, and one of the things that came out of it that I can say is non-public is that 2026 will be the year of the event, right? I think with the rise of AI, we're witnessing a commoditization of all digital communication. I always joke when I'm in front of large crowds of agents where I open up your Facebook or your LinkedIn, and you can tell that one friend who couldn't string a sentence together 12 months ago now is William Shakespeare with paragraphs of content. I think we're all developing a high-level filter for we discredit AI-generated content. I think there's going to be this drive and crave for human interaction.
We are going to operationalize scaling that through structured programs for people to host events, invite, and we are going to continue, at the corporate level, scaling as much of it as possible with a playbook.
Correct. Maybe switching gears a bit, maybe for Jesse, eXp has been very deliberately growing its international footprint, its commercial real estate footprint, and it's surpassed $100 million in revenue, and that's up 68% year- over- year. What is driving that momentum, and how repeatable is the playbook as you look to enter new markets?
Yeah, thanks. I'll answer the last question first. I think it's becoming more repeatable. We've developed the playbook. I was Head of International Finance for the last four and a half years, so I was very close to it from the very early days. There were lessons learned in the early days. International is a different animal, for sure. We've really honed down the playbook. Each new market is costing us less to enter. We're opening markets now with teams or 50+ agents, whereas before we were starting more small-ball. The playbook is getting pretty solid. Back to your first question on the financials, it's accelerating. The 68% year- over- year, it's actually the highest year-over-year growth rate. The bar is getting set higher and higher, and we continue to surpass it. It's coming from both.
We're over 5,000 agents, so still a lot of room to grow. International, 28 countries. We're growing the already launched markets, and there's still a lot of room to grow in those, even in our biggest markets, U.K., France, Australia, South Africa, Spain. There's a number of markets that are becoming fairly material for us compared to their historic performance, right? They're still very small in terms of the TAM in those markets. Of course, launching new countries. We'll have launched eight countries by the end of this year. We're doing twofold, growing the existing core international and then launching new markets.
To that point of about eight new countries by the end of the year, what strategic or economic thresholds need to be met before you launch in a new market?
Yeah, I can answer high level, Glenn. I don't know if you want to take any of it. The things that we look for in new markets is the compatibility with the model, regulation, obviously, if it's going to be difficult to enter. Every market is unique and different. The actual landscape of the real estate market is different. A big part of it is the leadership. We want to find good people that understand real estate, but also understand the entrepreneurial nature of eXp and that we are such a differentiated model. It's definitely a mix of things that we're looking for when we enter any new market. We're not just filtering on GDP per capita or something like that. There's a lot of subjective input.
Yeah, and I'll just add, we recently just launched in Ecuador, which is not a big market, but we brought over the number one producing RE/MAX team. And because of that, there was actually a large number of agents that have joined. And Ecuador is actually doing more volume than some of our more populous countries that we've been in business with in for quite some time. It is very leadership-oriented more than it is we're not looking as much at the TAM. It's very inexpensive to open up a new market for us. At one time, under sort of, we'll say, a previous regime, it was estimated that it would cost us somewhere between $500,000 and $800,000 to open. Now it's probably closer to like $100,000-$150,000 all in before we sort of reach some level of volume and break-even status. We brought our costs way down. It allows us to open up in markets that are fairly small but have good potential.
The company has been trying to expand its commercial footprint also internationally. You launched in the U.K. recently. What are the early data points, maybe agent count, your flow, cross-border transactions that you will track to gauge the success and scalability of entering those new markets?
I'll take that one. Commercial is a really interesting beast. In the U.S., you have three major players: CBRE, Cushman & Wakefield, and JLL. If you look at the total TAM, our research shows that there's probably 70,000 real estate professionals who sell commercial real estate. Those three big companies are normally in the urban cores, but something like 80% of the units. 80% of the volume is probably concentrated in those three companies, but the units is probably flipped on Pareto's law of 80/20. There's actually a pretty big long tail, call it sub-20 million. We think we're still early in that story where we've been able to cobble together a very compelling offer. For example, it is a national single brokerage. We're the only residential brokerage with a commercial arm that has an enterprise CoStar contract.
We can compete in a way that Keller Williams or RE/MAX, where it's fully franchised. Franchise by franchise, they have to go negotiate and don't have national buying power. Our commercial agents in the U.S. enjoy discounted CoStar, which is like the foundation of the commercial toolkit and a bunch of other products inside of the CoStar family, in addition to CRM and other products. As that's worked in the U.S., that's kind of the beauty of international, to Glenn's point, of CoStar started at $750,800 to launch, and now they've come down dramatically. That's because of perfecting the model. That's what we're doing with commercial, and we think that's a massive opportunity. Again, in the U.S., we enjoy a lot of infrastructure, both in residential and commercial. Internationally, it's more of the tooling and kind of the light version of what we're doing in the U.S..
Looking further down the line, the goal of 50,000 agents in 50 countries by 2030 seems ambitious, right? What are the biggest operational or just overall, generally, hurdles that you're facing?
A couple of things. One, if you look at a proxy in RE/MAX, there are 110 countries and territories, 50 is conservative, in my opinion. The arrival date, I think it's what's up for debate and w here I get excited is always on TAM. North America is about 1.6 in 2025 between the U.S. and Canada, 1.6 million agents. We enjoy roughly 4%, which is fantastic. If you look historically, RE/MAX, Keller Williams in their peak, were somewhere in the 10%-15% range in North America. We have an amazing runway in North America. When you look at international, in the top 100 countries, kind of the ones RE/MAX are in, some of our data, and it's not perfect because there's not licensing rules, but GPT and Google have somewhat said about 18 million professionals consider this real estate professionals, right?
That is an order of magnitude larger total addressable market. We try to balance size of opportunity, number of transactions, right? We just opened up Japan, which I am extremely excited about. I think we underappreciate the size of certain markets. The U.S. has about 94 million residential 1-4 single-family townhouse condo parcels that you can buy. Japan is 66 million, right? That is a massive opportunity. Where you have seen like in Anywhere, brand and Century 21, they have 950 offices in mainland Japan with that affiliate being publicly traded. When we say that is a kind of a green shoot, it feels like when Glenn launched the company 16 years ago, we are the only player. In other parts of Asia, where direct selling models flourish.
Maybe take a step back and looking at the broader landscape, you've said eXp is no longer the disruptor, but one of the scale incumbents.
I did say that. I've been quoted a lot. That was my Q3 comment that made a couple of headlines.
What differentiates your model most meaningfully versus both legacy brokerages and newer platforms that are entering?
Yeah, so I'll double down on Glenn's analogy on Netflix, right? Netflix started as a disruptor by mailing DVDs because it was competing with brick and mortar, with Blockbuster, pivoted really as the technology matured with high-speed internet. Netflix is now airing fights and sports, live TV, right? I think they've done a masterclass on pivoting correctly. The other company I'd like to make as akin to is Meta, right? They started as the disruptor to MySpace and then took over with Facebook, but then have bought the correct platforms. When they see competition, which is what I think our size, scale, capital has given us, is the ability to take a step back and go, "That's a great idea. We can either buy it, compete with it, build it." Instagram saw Snapchat show up, and stories came about.
TikTok showed up, and they came out with Reels. Our size and scale, and really the moat is the flywheel of the scale of the culture, right? Replicating the tech stack, especially in today's very quickly shifting AI world, is going to be easier and easier. To have a scaled enterprise where on one of our calls I disclosed, we sent 30,000 referrals from agent to agent across our platform. There are companies that do not do that many transactions in a year. We just did that internally. That gives us the ability, I think it is self-reduced responsibility to stay in curiosity and humble and analyze anything our competitors do and kind of do, "Does that scale? Does that make sense for our model?" Right?
I think the shift that's really bifurcating is there are the brand-first companies and Compass, Anywhere, Howard Hanna, Berkshire Hathaway, kind of that ilk, which is real estate-based, physical space, cap unsplit, and more support with a very expensive economic model. I think that tends to attract a solo producer that I would describe as a self-employed employee, and I would compare him to insurance salespeople, maybe a small practice doctor. We're tending to attract the entrepreneurial realtor. If you take our top 250 teams in 2024, they sold in excess of 75,000 homes. That would make us the seventh largest brokerage in the United States in that year. These folks have gross revenues in the $10 million-$20 million, right? They're not a $150 solo practitioner. They're business owners. We view ourselves as a platform that allows them to build whatever size dream they want and scale it, whether it's multi-county, multi-state, or multi-country.
We've seen some acceleration in broker M&A this year. How is eXp positioned to capture share going back to the attrition point as agents and teams seek some stability?
Yeah, no. A couple of things that history will show us. When this country starts to consolidate, it doesn't stop. If we go to the late 1800s, 118 railroads, like 10 years later, there was like four, right? Airlines, healthcare, it's happened over and over again. I think we're uniquely positioned to be a participant, right? We're not making any forward disclosures, but we are well-capitalized, asset-light, and in a position to take advantage of whatever opportunities the market bears for us. There will also be the organic pressure of that.
I can say that I've had more conversations post-September 22nd when the announcement was made from small independent brokerages who had no interest of ever joining a company, but now see that if we look down 3 years-5 years directionally, there's probably going to be a handful of players, 3-5 , which is what we've seen in other industries. There's no question we'll be one of them. It's just, are we in first or second position?
All right. From a regulatory standpoint, if clear cooperation or MLS policies were to change meaningfully, how could that impact your listing exposure and just your business generally?
Yeah, so I have strong, strong feelings about this. We are for transparency. I can't say it loudly enough. I do think that if we had to underwrite, that's a strong possibility. As of right now, we are the single largest. We are in pole position. We continue to focus to partner with folks that have dominant market share in their markets. From a scale technology standpoint, we are prepared for whatever version of the world looks like. We are uniquely already operating in that way outside of North America. That is how we operate in the U.K. That is how we operate in Australia. That is how we operate in South America and all over South America. We actually know how to do it. We think it's better for the consumer to have as much transparency as possible.
No matter which way the world goes, what we do know is over a 50-year period, 4 million-7 million Americans choose to sell their homes. We will be there to support them. Our goal is to have as least friction as possible. Just to be clear, I've said on many earnings calls, from a technology standpoint, if MLSs were to disappear tomorrow, we are fully capable of inputting listings, syndicating directly to portals, and bypassing every other mechanism necessary from a technology standpoint, from an infrastructure standpoint. We've been ready for this two years ago.
Right. Maybe let's turn now to just the housing market and micro backdrop. Many independents facing margin pressure and power market volatility. As you mentioned just now, industry consolidation shaping your recruiting funnel and your conversion rates.
Yeah, so it is hyper-competitive, right? Because there's less transactions. I think it's a bit counterintuitive because it feels even more competitive than in a normal market. I just think we're in the third year of a down cycle. There is pressure from folks who are hoping for a recovery. Fannie Mae telegraphed at the end of 2024 a 10% increase in transaction counts. That did not happen. We're barely touching.
That's the forecast.
That's the forecast. You got us once, you're not getting us twice. Jesse and I nerfed that all the way down to about 2%. We're just going to assume a current run rate and just underwrite on that because we've been able to maintain a fairly good model, right? We reported net income, not adjusted, net income last quarter, which I think we were the only one in our peer group to do so. That's become a superpower that we're leaning into.
Right. Speaking of activity and forecast today, how would you describe the housing market, housing activity, and to your point of agents, how is sentiment?
Yeah, so twofold. Real estate agents are some of the most resilient human beings I've ever encountered in my life. It is a...
You were one.
I am one. I still hold an active broker's license, which I think speaks to just the avatar the industry attracts. It is an entrepreneur. It's highly, highly aspirational. The sentiment is higher than most people would expect. That is partly because over the last three years, it's been so difficult that there's been a high level of attrition. We have seen of the agents we lost over the last couple of years, north of 60% of them, we lose to a job. They are not going to a competitor. They are not shopping for a better rate. They are actually giving up on their dream. What that means is twofold. The experts are actually taking market share. This is a really interesting fact. Year- over- year, we are up 2% in revenue disclosed. Our top 250 teams are up 15.8%. The bigger are getting bigger.
The richer are getting richer, kind of the classic taglines. From a consumer standpoint, I think it's extremely confusing because it is a hyper-local tale of a market. The suburbs of the Northeast, New York, Connecticut, where all of our listeners are probably sitting, have seen about a 3-5% appreciation rate. Historic rates are about 4.75% year- over- year over a 50-year period. This is sitting close to that, barely below average, but there. You zoom out and go to the southwest coast of Florida, like Cape Coral. Cape Coral is ground zero. It's about 11% year- over- year depreciation in price on the average. There are other markets that are softening. Miami was starting to soften. The New York election may give it a boost. That is to be determined. What it is, it's a tale of a local market.
The consumer, I think, is experiencing paralysis because we live in a hyper-macro information world. The headline is, "The real estate is melting down," and you could live in Greenwich, Connecticut, and that's not at all accurate. You could say, "The market's recovering," and you live in Cape Coral, and that's also not accurate. What we're seeing is a lot of disconnect between the seller and the buyer psyche. Sellers think it's 2021, and buyers think it's 2008. The answer is it's neither. We're educating and imploring our agents to be able to have tactical information at the local level. Meaning when you go to list a house, understand, is there four months of inventory, which makes it a seller market? Are you sitting at seven months of inventory, which is defined as a balanced market?
Are you sitting at 12 months of inventory and you're in a buyer's market? You got to be able to price that property below the comps to get ahead of a falling market, which I've described to agents as catching a falling knife when you're in a depreciated market. Again, knowing the difference, and not at the zip code level or at the county level, at the actual neighborhood level. The neighborhood is very different in a grouping of townhomes, single-family homes versus one condo tower. It is a market of experts at the moment.
To that point, despite a flat housing backdrop, right? Transactions per agent have gone up. What's driving that is team structure, tech, or just share gains from, like you said, the...
Yeah, it's a couple of things.
Inexperienced.
We are focused on the productive. And so we retain them at a higher level and the lower productive churn at a higher level. It is a flywheel effect that always pushes transactions up. If you look historically, when there are 7 million transactions, there is a larger denominator that gives us a lower PPP. When there are less transactions, a lower denominator gives us a higher PPP. Again, I always compare us to our peer group and just try to make sure that we're on the higher side of that.
Turning to outlook, what are you seeing in the signs of the market stabilizing? What are you seeing in transaction pipelines?
I made some bold predictions last year. I was cautiously optimistic. I'm back to pessimistically prepared. I would underwrite flat. I do think there is macro instability, whether it's conflict, GDP, debt, unemployment. What that translates to at the consumer level is it doesn't matter if you have a secure job. You read all this negativity. Your lease comes due. You have 90 days out. You may renew versus get into the buy-side psychology. I did a keynote at EXPCon where we gave them data. This is a fun fact that was disclosed at our presentation. We can send you the slides if you need them. Across multiple products, the average rent has now increased to about $2,030. It's an estimate. You can get the slide if you need the exact number.
On the same grouping of product, it was about a $60 variance to homeownership. That was the important one. A lot of people will say, "Maybe I'm going to wait three years." The danger is if you were to wait three years at average appreciation rates of 4.75%, you would have to wait for rates to come down all the way to almost 5%, about 5.18%, to get the same buying power with inflated dollars at the current inflation rates. I think there's a lot of people waiting, right? I was a realtor, as you mentioned. I sold close to 4,000 homes. People said, "When's the best time to buy?" I said, "1942, 1960.
You're saying there's a disconnect between sentiment and the underlying economics of buying a house today?
More importantly, it's your personal journey. I would say to a consumer, if you're not going to move for the next 10 years, your family ties are there. Historically speaking, over a 50-year period, you're better off buying, right? If you're not going to move because of your job, you're not on a government contract with an administration, you're getting moved out of the DC area. If your parents are down the street, you work at a factory, you work at a stable industry in that geography, the forced savings account that becomes homeownership, like the average net worth of a tenant is $4,700. The average net worth of a homeowner is north of $400,000, right? Being the only comparable difference on someone's balance sheet. The 30-year mortgage with a self-liquidating feature doesn't actually need account appreciation, right?
If you were to buy a house for $100,000, make the same monthly payments that you would have made to your landlord, at the end of 30 years, if the appreciation did not go up, you would have $100,000 in equity.
Right. I think people mostly focus on interest rates, but that's not.
No, that's actually I covered that in my talk. Historically, our interest rates are actually below a 30-year average, right? Your 30-year or 50-year period, the average is sitting at 7.7. We've had two anomalies in the last 20 years that are of historic proportion, right? The largest financial crisis since the Great Depression. We threw money out of a helicopter, per Ben Bernanke exact quote. You follow up with a 100-year pandemic. Both of those printed more money than we've had in the last, it was like 27% of all capital ever printed. I think there's an aberration in our memory of what rates should be. That doesn't mean we don't have an affordability crisis. It's just having to educate folks on the difference of expectation on rate to what is affordable and possible in your market.
To your point on affordability, what needs to change for that to improve? Because rates is not going to be.
There's a couple levers. There could be increase in wages, right? And then making sure folks are right-sizing to their opportunity. In certain markets, for example, we have a 10-year high of new construction inventory across the country right now. A new home built by a builder is offering opportunities as low as 4.99% on a 30-year fixed mortgage because they can do a buy-down rate. It goes back to the personal story and the personal journey and understanding what your goals are for you and your family and making sure you're making a proper decision, which should be with the advice of a qualified real estate professional, hopefully one that works with eXp.
Right. Maybe turning back a little to the financial performance. GAAP gross margin declined modestly due to more agents hitting their cap, which is good in and of itself, right? As productivity rises, how do you balance those incentive structures with maintaining sustainable margins for your shareholders?
I mean, I'll jump in. If you look at our peer group, our closest competitors also saw the same thing, right? Part of it is what do you do? What you do is you double down. You double down on expansion of market share and talent. Our thesis is everyone has the same pressure right now in a capped model. We are going to continue adding to that value stack. As it recovers to, again, a normal balanced market by any definition that you want to give it, we will be buoyed up by that. I'll let Jesse add to more specific on the financials.
I don't want to go after you, Leo, but. No, yeah, it's around six and a half. Our peer group that have similar cap models are actually far underperforming that margin. We have all the upside that Leo is talking about. It's that, is it a feature or a bug, right? When you have more productive agents that are capping, we consider it a feature. We want the productive agents. We want to win market share. That's how you win in the long term. Also, to hit on international. International, we have more favorable margins in the mid-teens. That's just due to the business models there. We can have more favorable splits and slightly higher caps just because it's open landscape in the markets that we're entering and competing in. There's upside as international continues to achieve scale, again, growing at 70% year- over- year. That margin will help our overall consolidated results.
Right. To the margin point, your transaction processing costs keep improving, declining 15% sequentially. Do you see more room for that to improve maybe coming from workflows? I won't put you over there. Getting more automated?
Yeah. I mean, I think we're living in unprecedented times, right? I think whatever analogy you want of the transformational power of AI, I've heard as much as fire to internet and high-speed internet, right? It's probably somewhere in between. It is transformative. Again, is there a hype bubble? Probably. If we look back to the dot-com bubble, the popping of that bubble gave us the discounted cost of fiber optic cable that wired the world, right? We have actually, because we're in a repetitive business model, been able to, I think, take advantage of those opportunities faster than other companies. We are constantly trying to level up and find efficiency, right?
Again, going back to my comment of the digitizing of communication, we actually think that as we achieve more margin improvement, it's not just going to go down to the bottom line. We'll be able to invest in the human-to-human experience, which will separate us. Because at the end of the day, real estate brokerage is a commoditized business model. The differentiation does come from the human experience and the scale and the community we can build.
G&A has also been coming down. Do you think that that will be sustainable as you keep scaling the business?
At this point, we believe that our baseline that you're seeing will be constant. Again, we want to get an uncorrelated growth between cost and revenue.
Right. Lastly, on the financials with stronger cash flow and what you just mentioned, a disciplined cost base, how is the company prioritizing capital deployment among AI, returning capital to shareholders, and maybe some M&A?
Yeah. We were just doing some math earlier today, and I think we've returned through buybacks and dividends over $800 million to shareholders, which is pretty crazy when you think about most of the companies in our space are returning money to bondholders and just debt repayment. We are kind of in a great position to just continue to sort of invest. I mean, you look at the management team that we have here, you look at the tools and technologies that we're rolling out, and we're going to continue to invest, but we're early innings into so many different things. International expansion, we feel like we're kind of at where we were in 2015 with domestic realty. And 2015 was the first year we doubled year- over- year in the domestic business. There were some catalysts to that.
You start looking at international and you see a lot of the same ingredients. Going back to the 50,000 agents in 2030, if we are entering that era where we're going to start to see the doubling of agent count and the revenues and everything else that goes along with it, we could go from 5 to 10 to 20 to 40 to 80 in fairly short order. It is not unreasonable to think that there's a lot of growth to unlock. We fully expect to add a number of additional countries next year. It may not be eight countries next year, but we're still going to add a lot more footprint, especially around areas where there's multiple markets that sort of connect to each other. We're really quite excited. I mean, we right now are running like we were in the United States.
We're running unopposed internationally. There's nobody internationally that has a similar business model of any note. That kind of gives us a lot of blue ocean to grow pretty quickly. Our value prop is finally getting to the point where agents are getting it. It took us about five, six years domestically before agents started to get our model. Once they got it, we took off. Internationally, we're right at about the same length of time. We're about five, six years into our international. Now agents, brokerages, business models, I mean, they're all recognizing that eXp is growing and we're growing fast.
Staying with the international, Glenn, that's been your focus for some time now. If you can share some of your experiences, the geographies where you are expanding, how they differ from one another.
Yeah. Where the U.S. and Canada are very similar markets, MLSs, boards, regulatory frameworks, governance, outside of North America, it's almost the exact opposite. There's virtually no regulation. Getting licenses, if there is a license, is about as easy as fogging a mirror. Reporting transactions in a lot of markets is spotty at best, meaning that there's a lot of countries where the agents regularly just do handshake transactions. We've got to put in more different types of infrastructure to account for, we'll just call it the common behaviors in markets to bring a more professional version of organized real estate to a lot of places. A big one is the way that portals operate internationally is totally different than domestically. Domestically, you can see pretty much every listing on anybody's website. Internationally, the portals are a big expense.
We think that our size is going to allow us to do some things that is going to give our agents, again, advantages. Part of it is because of some of the learnings that we've had here domestically, but some of it is because of some of our abilities to innovate and see cross-market what works and what does not work. International, I spent about close to a year working with the international team. I have shifted to Success on the personal development side of the house, but it is really in great hands. The team is running hard, opening up a lot of new markets. We have literally replaced, this is kind of an interesting sort of factoid, international because all the markets are disparate. There are so many different SaaS platforms where people were trying to sell us software for some pretty basic stuff.
Earlier this year, we literally, because of the power of AI, we've abandoned virtually every SaaS platform internationally, which means we're building all of our own websites. We're building all of our CRMs. We're building our tools. We're building our marketing platforms. We're building all of it in-house. We're generally doing it without software developers. We're doing it with subject matter experts, which is a lot of times you kind of learn at the edges what you eventually bring back to the main enterprise. We've learned a ton about what we can do at the edges, which I think is informing our future decisions at the enterprise level because I think that a lot of that stuff, I think we pretty much have a mandate that we will no longer sign a new SaaS contract. Any existing SaaS contracts, we hope will die a death and never have to pay for it again. There is a lot of stuff we are learning at the edges, but we are going to have a lot of fun the next few years.
Speaking of the next few years, as you look forward to 2026, what are the key milestones or proof points that will define the next stage of expansion for eXp?
At some level, we do not provide forward guidance, but I will tell you that one of the things that we look at is productive agent count. A lot of times we had historically, especially during our high growth stage, we focused on agent count for agent count's sake. Everybody measured themselves by agent count. What we learned when I jumped in on international, the very first thing that we did is said, "Hey, we have " 5,000 agents", but what is the real metric that matters?" It is agents that actually sell real estate in the quarter. We started to measure that.
That delta between actual productive agents and total agent count, those numbers are getting closer and closer to being the same number, which internally, that sort of turnover of productive agents pushes in a really good space to now think about we're at the point where doubling will start to actually show up in the agent count numbers. We did a lot of trimming of people who weren't paying us monthly fees, weren't turning in transactions, and we basically got rid of at least 80%-90% of that cohort. Now what we've got is agents who want to be here and are willing to actually invest in their business. Those metrics for us are kind of the metrics that we get really obsessed about. We still look at net promoter score. We use it a little bit differently than we did historically.
Historically, we just measured NPS, and we always wanted to go up and to the right. That's fine, but it misses an opportunity because if you're we now think about it as a sweet spot. We want to be generally between 70 and 80. We can be above that, but it misses an opportunity for efficiency improvements. Now we're using it kind of as a range that we want to be in. If we get too high, that's like, "Hey, there's got to be some things that we can do to automate, lower our cost, etc." What we've done is we went in, we've done that, and then our NPS starts creeping back up again, which is what we want to have happen. It's doing it in the idea of the G&A actually coming down over time based on sort of efficiencies. It is not so much that we are saying, "Let's cut costs. Let's cut costs when we know that we have room without hurting the agent experience." Those are some of the metrics we look at. I know, Jesse, you have some that you like.
You hit on them. The productive agents, I mean, we correlate so much of our financials directly correlate to productive agents. As Glenn mentioned, we noticed several years ago the bifurcation from just agent count, which is funny because still most of the industry compares us and competitors and everything else on agent count metrics. If you have a disconnect there, which as Glenn mentioned, we've greatly narrowed that gap, it shows up in your financials.
One last one. From an investor standpoint, what aspects of eXp's platform transformation do you think are misunderstood or underappreciated?
We're not a real estate brokerage. We're a technology company that just happens to sell a lot of real estate. We really are fundamentally built as a platform. Now we've got the best in the industry. You got Leo and Wendy here with me up here. We have the best, the best leadership in the industry. We're also the only company that's really positioned where every part of the stack is effectively a piece of technology. Technology right now is changing so rapidly with AI that we think that it's going to be pretty exciting probably by 2027. You're going to see so many improvements from just financial metrics because of all the things we can do that we couldn't do just 12 months ago.
All right. We'll wrap it up there. Glenn, Leo, Jesse, Wendy, Denise, thank you for being here with us. Thanks to everyone for joining us.