Morning, the Accelerant session, in case you don't know me, Adam Klobber, the insurance analyst at William Blair. For compliance disclaimers, please look on the website. Accelerant, I'll say two words, then turn it over to Jeff Radke, CEO and founder of the company. Accelerant is a really cool company, and I don't use that word lightly. probably the biggest trend in insurance in the next five ten years will be what i'll call it more hybrid models and accelerant is and that exaggerate is leading the way and i won't give you know spoil too much of jeff's talk but you really are seeing much more of a melding of distribution and underwriting capability and what i think is much more efficient and flexible nimble way in a business that really has defied those three factors and we've seen a lot of momentum again accelerant is leading the way here um and you know what we love it's an emerging business model it's grown a lot it's early and that's the time to us you want to look at these businesses so with that jeff thanks very much take on good morning everyone
so as adam said i'm jeff radke um i serve as the ceo of accelerant um this is going to be a lot more fun for me if you interrupt with questions i don't know why you care about how much fun this is for me but just there's a little fact um uh no please do interrupt i know generally people don't but uh i i think i think uh you're if you've got if i didn't explain something well probably one of your fellow investors is also wondering the same thing so please do feel free to interrupt um i'm going to start by telling you a fib um but it's a useful fib um as if you start up thinking about accelerant as a pure generalist and a mental model that wouldn't be terrible to have in your head is uber and here's what i mean by that uh before uber maybe we didn't know this but we had thousands and thousands of people that needed rides right and we had hundreds and hundreds of people that wanted to make money driving people around and they never connected right they never ever connected uh until uber and uber is a toll gate right uh on that flow of funds that's not a terrible analogy to start with for accelerant okay um so accelerant uh i suppose i I should use it. Accelerant is a specialty insurance marketplace, and what we try to do is we are the connector between underwriters, insurance underwriters, that don't work for insurance companies. We call them MGAs or MGUs, doesn't matter. They're underwriters that hung out their own shingle. We connect them with risk capital. That's what they need to be able to issue policies in as efficient a way as possible. And we get paid by risk capital, a percentage of premium, about 8%, for doing that. So we get paid by risk capital to source, manage, and monitor that portfolio specialty business. So with that, we'll start to kick it off. When we started in 2018, we had three hypotheses. The first one was that these independent MGA underwriters were going to out-compete the monolithic insurance companies every single time. And boy, has that proven to be right. Adam, I don't know, the MGA market has grown by like 70% or something. Leaps and bounds. Yeah, gigantic, right? And it's just simply out-competing the big monolithic insurance companies. companies as someone who worked in many of those monolithic insurance companies it's like the smallest surprise ever right you could have predicted it from a mile away that uh a a owner operator would beat that monolithic insurance company no matter what the endeavor was uh the second one is that modern technology is what's going to pop is what's going to power the exchange the corollary that i don't have the courage to write down is that no this is bold no big legacy insurance company could possibly deliver that technology i know that sounds crazy because they have tons of money they have plenty of money but again having worked inside them it's just the system is rigged there's no possible way i was the chief operating officer of the specialty insurance company back when, this dates myself, back when you had a server room. And I walked into the server room and way in back, you guys remember Windows NT? Remember the computers were beige from IBM? And I pointed and I said, what is that? And they were like, don't even look at that. That's our surety business. The guy who wrote it is gone. No one knows how to fix it. Don't even look. No one's allowed to touch it, right? That's sort of not an atypical story about the industry. So disaggregated specialist underwriters are going to win. Modern technology is the way this has to go. And then the third one is that I knew from my time on working as a reinsurer and as a specialty insurer, I knew that what risk capital really needed was a source of really well-performing, highly diversified, low volatility business. When I started in the 80s, when I started late, but in the 80s, when I started in the late 80s, about half of your book as a reinsurer was high risk stuff and half of your book was low risk, lower return, and it was sort of like a barbell, like investing. Fast forward 25 years, 30 years, what had changed, the insurance companies had gotten so big that they didn't need to see that low volatility stuff. So the only thing that they shared was the high volatility stuff. That makes that low volatility portfolio business that Accelerant has, $4.2 billion of it, incredibly valuable. Did I explain that well? Does that make sense? So those were the three. If you want to draw a picture, we've got the supply side. On the left side, we've got these MGAs, Managing General Agents is what it stands for. There are 296 of these companies that we serve as of the last quarter in, I think, 22 countries, EU, UK, Canada, and the US. They are the supply side. So they have the distribution relationships and they have the underwriting expertise to to be good risk pickers, sort of like stock pickers, I guess, is an analogy. So what we provide to those MGAs, to what we call members, what we provide to them is all sorts of support services, which we'll get into. And what they provide to us is a high quality portfolio of risk that's closely managed by us. And the reason Accelerant can closely manage it for the first time in my career is because of the technology. I don't know how many of you guys are parents and how many of you parents have Life 360 to spy on your kids. The great thing is you don't have to nag them quite as much as before, right? Because the machine spies on them. Well, same thing here, right? We don't have to nag our members because the platform spies on what they're writing. We know what they wrote last week. We know what they wrote the week before. We know at what price. We know how risky the business is. So all this stuff about MGAs, it's a black box. Yeah, if you didn't have the technology, that's how I ran my whole career up until Accelerant. You didn't know. Now we know. And on the other side of the platform, you have the demand side. So we call it risk capital. That's made up of insurance companies, reinsurance companies, and institutional investors. That mix has changed over time. We started with reinsurers. They'll always be a cornerstone. We've got great relationships, really big relationships with great reinsurance organizations like Lloyd's of London, like Allianz Re, like Hanover Re, like QBE ReScore. So they're all very, very big relationships with experts. Why do they pay us 8% to source, manage, and monitor the portfolio? Because they can't do it. They don't have the technology. And probably they don't want to do it. We'll get to this in a second. But our portfolio business, when I say low volatility, I just want to underscore that the average premium per policy in the U.S. is just over five grand. So, you know, people say SME, small to medium-sized enterprises, it's like teeny is our average, tiny, tiny, tiny. These big organizations aren't good at collecting a whole portfolio of that stuff. And then there we are in the middle, right, as the platform. And what does the platform have to do? And you can see it there. You have to apply human judgment and experience that's turbocharged by the technology. So in other words, that smart, experienced woman or man shows up at just the right time because of the technology. So our underwriters never have an easy day. They just deal with problem case, problem case, problem case, because the platform serves it up one after the other. Way more efficient and way more effective. I'm going to skip that in the interest of time. And I think this slide has got a lot of numbers on it. I hope you can see it. I'll be careful to read them out for you. I'm also a little colorblind. what is that green color teal is that right yeah exchange written premium what color would you call that green you and i get along great right there's eight colors no right um uh the green one exchange written premium that measures the premium that flows through the platform and you can see that that's grown very very nicely up to 4.2 billion um i can't see from here what that growth rate is what is it 20 24 percent in the in the last sort of comparator so growing really really well uh because accelerant if you're an mga being an accelerant member is the way to maximize your success and that's a really important statement right our net promoter score is in the mid 80s to give you a comparison apple i think is in the 40s um so to have an insurance organization be that highly regarded is very, very atypical. So what's important here? Really great growth and really great margins and the EBITDA margins are going to get better and better and better. We feel very, very comfortable pointing you towards 50 and moving up. And the reason we say 50 and moving up is because the mix of our business is moving more and more towards just the risk exchange and less underwriting. We have a business where we own insurance companies to facilitate the transfer of that business to reinsurance companies. That's becoming a smaller and smaller part of our business. So the EBITDA margin is going to get better and better. So to give you, here, these numbers are a little bit bigger. 296 MGAs, 4.3 billion dollars on a trailing 12-month. It's up 24%. And the key here is the market that we're playing in, the addressable market, is over $250 billion in the territories that we're currently in for the classes of business we currently write. Gigantic. So people ask, geez, you keep saying each quarter that your pipeline's bigger than it's ever been. When are you going to reach the bottom? How many MGAs are there out there to serve? And what we say is, I think that's the wrong question. We form MGAs and MGAs are formed away from us all the time, right? This is a secular, this is a fundamental shift that we're seeing in the market. So our addressable market is that whole 250 billion. I guess the point here is we haven't, we've just gotten started is the way I'd describe it. Rates. One of the things that, if we're fortunate enough to have interested you about the company, one of the things as you start to read is a concern, well, it's not a concern, it's a truth, that insurance rates have been decreasing over the past, let's just say 12 months, to keep it simple. And people are worried, quite rightly, that that's going to compress the profit margins of insurance and reinsurance companies. That's true. Why do you think, Jeff, that Accelerant will still be able to have that 8% take rate? And I guess part of the answer is here. I told you that we have incredibly small groups of policies on average, right? $5,000 per policy in terms of premium. Those small policies are not subject to the rate swings up or down that larger policies are. I remember, I think, we had a really big quarter where rates were up 14%, whereas if you were reading the industry rags, they were talking about property being up 40%, and our best ever was 14%. I think last quarter we were up one, but that's better than down whatever the rest of the whole market is. So keep in mind that our portfolio, because it's small business, is not subject to those big rate changes. So much of the business is what we call minimum premium, which is, hey, if you want a policy, it's going to cost you X. So it doesn't really matter where the rate levels are. What that means is that we're highly, highly confident that our relative profitability is increasing relative to all the other business that our risk capital takes on. Does that make sense? So it's to a certain extent, it's a relativity game. So if you think about Accelerant, Accelerant's business consumes a very, very small amount of capital. It's easier for me if we just pick one. Hanover Rhee. Hanover Rhee writes all kinds of business, billions and billions and billions of dollars of premium. And what they're really, really worried about is seismic activity in Italy, seismic activity, earthquakes, being all fancy, earthquakes in Italy, earthquakes in California, but especially hurricanes on the East Coast in Texas. That's their big exposure. So when they bring on hundreds of millions of dollars of premium from Accelerant, the correlation is zero to their big risks. So their capital allocation model returns a tiny number. So if we're making about 8% of premium as their profit margin, that's an extraordinary return on capital. In relative terms, there's no other business that they write that returns that higher rate of return on capital. And we've been told that by Hanover, Allianz, and QBE. We didn't need to be told that. I've worked, we've worked in companies like this we know how the math works but it's really really powerful in a softening market accelerant and its portfolio business is ever more valuable so one of the things this slide forget it one of the things that we talked about was real difference makers technology and i can only imagine uh how tiresome it must be as investors to hear AI, AI, AI. Every once in a while someone will say agentic for a little spice of difference. Unfortunately, I can't avoid it. So what's really important, and when people talk about data, I think there's this innate skepticism which I don't blame you for having. Let me try and make it sort of tangible and real. If you license a normal policy administration system as an insurance company which i think they all are essentially all of them do there are boxes that you get to fill in about this risk so uh what class of business we'll do property right so a property risk comes in and it's an office building and it's an office building in kansas city and you get to type in a bunch of stuff about the risk uh in your policy administration system that policy Property administration system was created when you used to get yelled at by the IT department for how much data you were storing, because storage was expensive, right? So there are about eight fields on average that you can type in about that property risk. So when you type in the zip code, now you've got seven. You think about it, think about it, okay, so how about limit? That seems important, right, okay, now six. I guess the point is, insurance companies store almost nothing. Now to make this make sense, the underwriter has got an Excel spreadsheet where he's typing in all the other stuff that matters. What's the crime score? What's the fire suppression system? What's the fire protection code? I'm nerding out on you. A whole bunch of insurance nerd stuff, right? And all that goes into a spreadsheet and it's lost forever. So the difference that Accelerant started with because we knew the problems is we started with the snowflake layer. There's nothing earth shattering about that. In the industry, it's pretty unusual. So we store 200-ish on average attributes about each and every risk, each and every class of business. And as you start to issue 156, I can't remember what it is, million policies, you get a long list. The rows aren't as important as the columns because the columns are the attributes of the risk. And over time, you keep building and building and building. That's what the machine learning models feast on. And then when you lay AI on top of the machine learning models, all that just happens way faster. So we're to the point where essentially all of our premium, 80% of our premium, will be covered by machine learning risk models by the end of the year. And what happens then? What you can do is you can then hyper segment. And I think theoretically this is important. An insurance rate is an average of a cohort of what's supposed to be homogeneous risks. So a whole bunch of apartment buildings will get the same rate. But you know that average is wrong. You know that half of them, that rate is too high and half of them it's too low. What happens if I told you life is as simple as figuring out which is which? It's not that hard if you have the data. It's not that hard. And just to be clear, I'm not talking about like the human genome or anything like that. I'm talking about much simpler stuff. Like if you're a pub, if you're a restaurant or a bar, if the bathrooms are on a different floor than the main eating floor, you're worse because people fall. so simple, right? But if you don't ask, you don't know, right? Another one, all the good ones come from the UK. Another one, in Liverpool, if you insure a pub that's north of the Mersey River, that's the bad one. South is good. Anyone who knows the football, so the scrappy team is supported by the people on average who live north of the river, so the fights in the pubs on Saturday cause all sorts of damage and liability claims. There are many other examples, but what I'm trying to do is it's not like magic. It's pretty simple stuff that you can drive out. And did I know any of this? Did we know any of this? No, right? We needed the machine learning models to say, hey, I found something interesting, right? Or machine learning models that now point us towards claims where we think we have the highest probability of being able to recover from some other responsible party that's brought our loss ratio down by a point that's huge and what a difference so technology is making all the difference inside this platform right the human stuff is good but we've got a lot of smart people there are a lot of smart people out in the world that work at many different places that technology platform and the combination are really something else so if you want to know how we make money uh this is the slide if you're on a desert island and someone wanted to know about how accelerant makes money this is the one um so we've got three businesses uh the eight percent that i told you about exchange services of where the the premium flows through and it's a toll gate eight percent times 4.3 is about 360 million bucks. We own a minority of the MGAs that we support. If you want, I should probably be a little bit more precise. We have many, many minority interests, and we have some wholly owned, but they're on average, I think there are 46 of them compared to the 296 total MGAs. So most of the MGAs we don't have ownership in. However, where we have an ownership interest their net commission is 18 percent and they write 1.3 of the 4.3 so that generates another 219 million and then finally we've got these insurance companies that facilitate the transfer of the business to the reinsurers we're using that less and less and using other people's insurance companies more and more but where we use our insurance company we collect expect about 3%. We tell people to expect zero. It's normally been positive, but that's not our objective. This is also the part where I say, please, please, please pay attention to this. We are not seeking to maximize revenue. We're seeking to maximize fee-based revenue. The first two boxes, we want to be as big as possible, and that third box, we want it to be as small as possible. It's profitable business, but it consumes capital, and we We want to be capital light. We want to be that exchange that we were talking about. Why do we have it? Is that the question? The question was why do you even have it if you don't want it? In the beginning, we had to show the industry how it worked. So we had to have insurance companies that would say yes so that we could give the customer service we needed to the MGAs. And then we could now take that business to those reinsurance companies I mentioned. What's happened once it's proven, right, it's always the second fund is always easier to raise. Now that we're in the second or third fund, we're able to get companies like Lloyd's of London, QBE, Axis, they're happy to be the insurance companies in that mix. Now, they wouldn't have been originally. Thanks for the question. So what's the growth algorithm? Again, pretty small, pretty hard to see because it's small, but this is the cohort graph, sort of the classic cohort graph. Eighty-ish percent of our growth every year is from existing members. The net revenue retention, I guess, is about 122%. That's the engine of growth. And then, of course, we add new members every year at a slightly increasing rate. Eventually, they're going to become important, but never in the first year. It takes them a year to sort of organize themselves and for us to organize ourselves with them, for them to really start taking off. Now, how do they take off? How is this growth occurring where it's not suicidal? Because if any of you are not generalists, if you're insurance specialists, growth rates like this spell trouble. And the reason is, and this is so important, the reason is we're not growing on a policy by policy by policy basis where we have to compete and win for new business. These are all book roles, right? Every single, when an MGA switches a product to Accelerant, it moves the whole thing. There's no competition. There's no new business sort of discount that you have to give. We know what the performance of that portfolio is. And so as it moves over, we can do so with a high degree of confidence. So that's the growth engine. I guess the way I'd sum it up is an organic growth engine, capital light headed towards even capital lightness, incredibly attractive and growing EBITDA margins, and that growth rate into a total addressable market of like $250 billion, we don't feel the constraints, right? If I told you what my long-term expectations were, my two colleagues at Feint, but they're dramatically different than $4 billion. $4 billion is we're just getting started. So I probably went over a little bit, but any questions? What's creating more and more MGAs? Boy, you wouldn't ask that question if you ever worked at an insurance company. It is horrible. You don't make any money to speak of. You have no control over your own destiny. You're forced to have terrible technology. On average, this is right. So any specialty underwriter that has proven that has a real reputation and great distribution relationships can monetize that expertise and relationships. Dramatically different economics if you jump out and either join an MGA or start an MGA. So in a soft market, what insurance companies do is they cancel their MGAs. First, they start talking bad about MGAs, and that is about two earnings call cycles, two quarters. No, they do. They have to talk badly about them for two quarters. Then they cancel them, and when their top line goes down or their growth slows, They say, remember, I told you last quarter, MGAs were going to have to get trimmed back. So if I worked at a traditional insurance company, I'd be cutting MGAs because I don't know what they're writing. I don't have the ability to watch, right? Old, experienced folks at MGAs know that that's what's happening. So that's why our pipeline is getting so much bigger. We have MGAs that know that what Lloyd's is about to do and what the traditional market's about to do, which is start trimming MGAs. And they're coming to us because they know that we're a stable source of capacity. So that's the MGA demand, right, or the MGA supply, I guess. On the demand side, if you think about a soft market, our business looks better because it's this small, it's not as affected by rate. So our business looks in comparative terms better. So we're seeing increased demand from the reinsurers and the institutional investors. And this is all underpinned by the fact that we're not geniuses, nor are we alchemists. The business that we write is not subject to these big rate swings, the small, small stuff.
there's a question. Okay. Well, this is coming in. We'll have a really good breakout. And Jenny, so please join us. But thank you, Jeff. Thank you.