Accelerant Holdings (ARX) Q4 2025 Earnings Report, Transcript and Summary

Hello, everyone, and welcome to Accelerant Fourth Quarter and Full Year 2025 Earnings Call. Please note that this call is being recorded. [Operator Instructions]. I'd now like to hand the call over to [ Hamed Ali Yaqorbi ], Investor Relations at Blueshirt Group. Please go ahead.

Good morning, everyone. Welcome to Accelerant's Fourth Quarter 2025 Earnings Conference Call. Joining me today to share our results are Jeff Radke, Accelerant's Co-Founder and CEO; Jay Green, Accelerant's CFO; and Ryan Schiller, Accelerant's Head of Strategy. Their remarks will be followed by a Q&A session. We issued a press release and presentation concerning our financial results for the fourth quarter of 2025 yesterday, and they are available on our Investor Relations website. Before we get started, I would like to remind you that our remarks today will include forward-looking statements, including those regarding our future plans, objectives, expected performance and in particular, our guidance for 2026. Actual results may vary materially from today's statements. Information concerning risks, uncertainties and other factors that could cause these results to differ is included in our SEC filings including those stated in the Risk Factors section of our filings with the SEC. These forward-looking statements represent our outlook only as of the date of this call. We undertake no obligation to revise or update any forward-looking statements. Additionally, the matters we will discuss today will include both GAAP and non-GAAP financial measures related to both our consolidated results as well as our operating segments. Reconciliation of any non-GAAP financial measures to the most directly comparable GAAP measures is set forth in our earnings release. Non-GAAP financial measures should be considered in addition to, not as a substitute for GAAP measures. Finally, today's conference call is being recorded and webcast. Now I will turn the call over to Jeff.

Good morning. Thank you for joining us to discuss our fourth quarter results. We had a fantastic quarter, beating our expectations on exchange written premium, third-party premium and adjusted EBITDA. All of these figures were in line with our pre-released results. Like last quarter, in addition to our earnings press release, we provided a presentation with more detail on our company. We also filed our first Form 10-K last night. All of these materials can be found on our Investor Relations website. Let me start with the topic that is on many of your minds, Artificial Intelligence. AI is the architecture of our business. It is embedded in how we operate and how we win. From our founding, we designed our platform to capture, structure and continuously learn from specialty risk data at scale. And I'll be direct, the more AI transforms the insurance industry, the more it plays to our strength. We've built what we believe is the largest proprietary decision-ready data set in specialty commercial insurance. We use that data to power algorithm-aided underwriting at scale. And we distribute that advantage through the most technologically adaptive segment of the market, MGA. Everything we are about to describe flows from 3 structural advantages: first, our proprietary usable data. Next, algorithm-supported underwriting and last, technologically adaptive distribution. Let's start with the first. Our moat is our proprietary decision-ready data. We have assembled 134 million rows of specialty insurance data across 58,000 unique attributes. All of it is proprietary. But data alone is not the differentiator, usable is the critical word. Our position in the value chain allows us to stitch together upfront underwriting submission data with the claims outcomes on the back end. That seems obvious, but it is rare for one entity to have this ability. That closed feedback loop, submission to binding to loss emergence is what transforms raw information into decision-grade intelligence. We don't think anyone else can do that across a multitude of independent MGAs and 600-plus different specialty products. We then augment this proprietary data with third-party data to further enhance our understanding of risks. The combination is powerful. The internal performance linked data layered with external enrichment. No large language model or AI start-up is going to have that kind of data, and we're adding more and more each day. The difficulty of building this kind of infrastructure should not be underestimated. Industry modernization efforts have struggled to connect and synthesize data across multiple different systems. Our platform was architected from inception to do exactly that. In fact, 34% of our workforce consists of engineers, data scientists, product managers and designers. Experts that are focused on building and honing our platform. And as AI capabilities increase, the value of our data set compounds. In a world where models will be increasingly commoditized, differentiated data is the durable competitive advantage. And every submission, every underwriting decision, every claim flowing through the Risk Exchange deepens that advantage. Our second structural advantage is our engine. Data alone does not create value. It must inform decisions. The set of tools that we have developed mutually support one another to form an end-to-end engine. We deploy risk scoring models, AI risk ingestion and real-time portfolio monitoring to optimize underwriting performance. As an example, just last month, we uploaded a portfolio of 1,679 new risks and in just over a minute, our engine told us that 84% of those risks were in appetite. 9% needed underwriting review given specific concerns identified by the tool. And 7% were simply out of appetite. The engine then went on to draft exclusions and coverage specifications to optimize underwriting performance. That achievement is monumental. That would have taken a team of people at least a week, and that's how we continue to win in the market at increasing scale. Our engine provides tangible advantages for Accelerant and its members in 3 ways: upfront underwriting precision, real-time portfolio management and claims optimization. Let's dive into specific examples of each. Upfront underwriting precision. We identify complex, nonobvious relationships across variables like financial health, industry stresses, crime indices, prior loss patterns, and we integrate all those insights directly into the member underwriting workflows. That enables more precise risk selection and pricing. It also helps us to systematically avoid loss driving outliers. That is a powerful combination. Proprietary, data-driven, AI-enabled tools pushing profitable growth, while avoiding poorly performing micro segments. That is how, based on our onboarding data, after the risk scoring models have done their work, members see a 2- to 3-point average improvement in their gross loss ratio, while growing at a rate over 35% since inception. The second advantage is real-time portfolio management. We ingest member data as it is created and monitor performance continuously. That allows us to detect deterioration earlier, protect underwriting margin and identify attractive growth opportunities. That's how we maintain a loss ratio in the low 50s. Finally, claims cost optimization. Our predictive analytics prioritize claims handling and surface subrogation opportunities, reducing loss creep and preserving profitability. Since inception, we saved our risk capital partners at least $100 million of loss using this capability. In total, the trajectory is clear. Underwriting is becoming increasingly autonomous. We've been building towards that shift for years. Whether underwriting is executed by members using our tools or as is happening increasingly automated within the platform itself. The Risk Exchange remains the infrastructure that evaluates, prices and aligns capital. Strategically, that means that Accelerant wins regardless of who originates the business. Every policy bound and every claim resolved strengthens the model and improve the next decision. Our third structural advantage is our technologically adaptive decentralized MGA. From the very beginning, we have focused exclusively on MGAs because they are more agile and faster at adopting technology than traditional insurance company incumbents. And AI enhances that advantage. Members on our platform gained access to capabilities that would be difficult and expensive for them to build independently. With our help, they achieve faster submission triage and quoting, AI-driven coverage recommendations, enhanced risk selection tools, real-time portfolio insights and life cycle management, policy to claim. Today, our members can ask Accelerant AI what's driving my loss ratio? And they can track the tools reasoning to identify key performance drivers and unlock unseen opportunities. MGAs are entrepreneurial specialists operating in defined niches. That means that when Accelerant empowers its members with improvements in underwriting precision and faster response time, it translates directly into market share gains. And anything we can do to make our members better, makes us better. So what does this all mean practically for our business? First, we can do diligence potential new members more quickly and completely. Our tools enable us to analyze their books of business and design underwriting guidelines in 20% of the time that it would take us to do so manually. The move from prospect to producing member takes half the time. When we onboard over 50 members a year, that makes a big difference. Second, once onboarded, those members grow more quickly with 25% underwriter productivity uplift emerging in our initial efforts. Third, we can lower loss ratios for our risk capital partners, delivering more predictable underwriting performance as we grow. That is highlighted by our attractive loss ratio track record. The future of specialty insurance will be data-rich, near autonomous and capital efficient. AI enables Accelerant specialized underwriters to win even more in their markets and moves capital closer to origination. What's the impact of all this? When data, underwriting intelligence and capital alignment are integrated into a single infrastructure layer, that layer becomes disproportionately valuable. Our Exchange Services segment already reflects the operating leverage of that architecture with near 70% EBITDA margins driven by near 0 marginal cost on incremental data and recurring revenue economics. And as automation increases, those structural advantages become more pronounced. Before leaving the topic of artificial intelligence, let's take a step back. At Accelerant, we set out to make the specialty insurance market work better for everyone. We believe the Risk Exchange is fulfilling that mission more quickly and completely because of the 3 structural advantages AI gives us. Turning back to the quarter. For those newer to our story, the Risk Exchange is a two-sided platform: the supply side and the demand side. The supply side is driven by the specialty underwriters, our members, who underwrite specialty insurance policies. These policies fuel our exchange. The demand side consists of our risk capital partners comprised of insurance companies, reinsurance companies and institutional investors. Those risk capital partners pay us a fee for access to our portfolio of policies. We sit in the middle, sourcing members, monitoring them and helping them to grow profitably by leveraging our AI advantages I just described. We leverage the scale of our Accelerant platform to deliver technology to our members that they would struggle to develop on their own. As discussed, our engineers and data scientists are developing agentic solutions for our members and for internal use in the Risk Exchange. These solutions help members grow more quickly, reduce their loss ratios and thus increase their profits. Our members value these tools and services, which is why we've consistently had an industry-leading 80s-plus Net Promoter Score. And Accelerant's reputation as the preferred partner for the best MGAs in the world is what drives the growth in new member MGAs joining quarter after quarter. For the demand side, we create value by sourcing, managing and monitoring a high-quality portfolio of insurance policies. Why is our portfolio of business so valuable to our risk capital partners? Well, it's stable, diversified and highly profitable. You can best measure that using gross loss ratio, which has been in the low 50s. That attractive loss ratio leads to consistent and attractive returns for our risk capital partners. With that overview complete, let's talk about the KPIs that we saw this quarter. Now last quarter, I set out the 6 key metrics that track the health of our business. During the fourth quarter, all 6 of those metrics were better than we forecasted. On the supply side, the first KPI is exchange written premium. Exchange written premium was $1.1 billion for the quarter. That's 24% year-over-year growth. Recall, we had a large premium, low-margin member that we terminated in Q3, excluding that 1 member, the year-over-year growth would have been 32%. The second KPI is our member count, and we continue to attract the best MGAs in the world with 15 additions in the fourth quarter, bringing the total to 280 as of year-end 2025. And the third KPI for the supply side is net revenue retention. For us, net revenue retention is the trailing 12-month exchange written premium growth of our pre-existing members year-over-year. That includes terminated members. Net revenue retention was 126% for the quarter and 131%, excluding the large premium, low-margin member we terminated. So our members continue to benefit from the advantages our proprietary data and tools give them. On the demand side of the platform, our first KPI is gross loss ratio. That measures how profitable our portfolio is for our risk capital partners. We target a low 50s loss ratio over time. And for the fourth quarter of 2025, the gross loss ratio was 51%. The second demand side KPI is the amount of third-party direct written premium. That measures our ability to attract insurers to participate on our Risk Exchange. Over the medium term, we expect this measure to reach 2/3 of total exchange written premium. And we have made stellar progress on this measure in 2025. For the first quarter, the measure was 19%. In the second, 27%. Third quarter, 32%. And for the fourth quarter, 40% of exchange written premium. We are well on our way to our medium-term goal and have demonstrated our ability to attract a growing and diverse stable of insurance company partners. The third KPI on the demand side is our net retention. That's defined as a trailing 12-month ratio of premiums retained on Accelerant's balance sheet to the total exchange written premium. One important reminder here, by lowering our net retention, we also lowered the revenue and EBITDA we book in our underwriting segment. Now we welcome reducing the revenue in our underwriting segment as it means we have placed more risk with our risk capital partners and generated relatively more fee-based revenue. For the fourth quarter, our net retention was 9%. That's in line with our expectations. In total, those 6 metrics, 3 on the supply side and 3 on the demand side, capture the health of our business. Short answer, we're executing extremely well. And because we are executing so well, we exceeded our guidance on all metrics. Ryan and Jay will share more details. Ryan?

Thank you, Jeff. Starting on the supply side. As Jeff said, in Q4, we delivered $1.1 billion of exchange written premium, a 24% increase from last year. And that number was 32%, excluding the terminated low-margin Canadian member we mentioned last quarter. For the full year, we produced $4.2 billion of exchange written premium, representing 35% year-over-year growth or 41% excluding that same member. Either way, really strong growth. That expansion came from existing members boosting new volume on existing products, new products written with us and, to a lesser extent, rate. Existing members represented 75% of our growth in 2025 and more than 80% of that comes from volume. While rate continues to be additive to our top line, it was just 3 percentage points of our 35% growth in 2025. Volume has always been the core driver of our growth. As a reminder, our book of business is 95% policies that are under $10,000 in annual premium. On these small premium policies, you don't typically see large rate acceleration or deceleration. So Accelerant's results are not as meaningfully impacted by the insurance pricing cycle. And as a side note, we are very comfortable our rate is keeping up with loss trend. The balance of our expansion comes from new members we add. In 2025, our number of members grew from 217 last year to 280, 63 additions, which is our best annual result ever and represents 29% year-over-year growth. We believe our member count is a good leading indicator for future exchange written premium. At the end of Q4, we had over $4 billion of annualized premium in our member pipeline, another record amount giving us confidence in our future growth. In 2025, we made great strides on the captive front. We added over $40 million of captives premium to the Risk Exchange, which is a testament to the team building out an entirely new, high-quality offering. We are just scratching the surface there and believe over time, captives could be a multibillion-dollar opportunity. Over the long term, what we're focused on being is a predictable organic growth machine as we expand into a really big market. Moving to the demand side. We ended the quarter with 95 risk capital partners, up 3 from last quarter, all 3 are Risk Exchange insurers. That brings our Risk Exchange insurer count to 18 versus 13 at the end of 2024. Our goal is to maintain a diverse group of risk capital partners to maximize the stability and efficiency of our platform in the long run. We seek to use third-party insurance company for a growing proportion of the portfolio over time, which in turn improves Accelerant's capital lightness. Our medium-term expectation is that third-party insurance companies will represent approximately 2/3 of exchange written premiums. And as you heard from Jeff, we have made great progress here with 40% of our fourth quarter exchange written premium coming from third-party insurance companies. And we've also made great progress in diversifying our portfolio of third-party Risk Exchange insurers. The largest Hadron has been declining steadily throughout 2025, from 67% in Q1 to 58% in Q2 to 54% in Q3 and finally 47% in Q4. When we do use Accelerant owned insurance companies to issue policies, we transfer the vast majority of the underwriting economics to our reinsurer and institutional investor risk capital partners. In 2025, 9% of the premium written was retained by Accelerant insurance companies, driven by regulatory minimum retention. We continue to expect net retention to approximate 10% in calendar year 2026. To summarize, our goal is to preserve a diverse group of risk capital partners to maximize the stability and efficiency of the platform for the long run. And now I'll pass the baton to Jay, our CFO, to address our financial performance.

Thanks, Ryan. Hello, everyone. I'm excited to share our fourth quarter performance and guidance for the first quarter and full year 2020. As you've heard this morning, we had a great fourth quarter with continued strong profitable growth and expanding margins. We placed $1.1 billion of exchange written premium, an increase of 24% year-over-year and 32%, excluding the 1 member that Ryan and Jeff both mentioned. That resulted in revenue of $248 million, which grew 30% year-over-year. As you have consistently heard from us, our goal is to maximize the revenue associated with our Exchange Services and MGA Operations segment, while continuing to keep our Underwriting net retention near our 10% target. You will see from our segment results for the quarter that we are successfully executing on that strategy. Adjusted EBITDA was $71 million, growing 52% year-over-year. Adjusted EBITDA margin rose to 28%, up from 24% last year. Included in adjusted EBITDA was more than $2 million of nonrecurring investment gains from our stake in the same own member that concluded a third-party capital raise last quarter. Excluding that investment gain this quarter, we generated EBITDA of approximately $68 million with excellent performance in each of our operating segments. This represents an increase of 132% over last year when you exclude $17 million of investment gains from fourth quarter '24 results. This momentum is driven by strong top line premium growth and consistent operating leverage improvement. Adjusted net income for the fourth quarter '25 grew nicely to $51 million compared to $39 million last year resulting in $0.23 of adjusted net income per share. We measure adjusted net income as GAAP income less profit interest distribution expenses, share-based compensation expenses and other nonoperating expenses net of tax. In the fourth quarter, you will also see $19 million of profit sharing expenses related to Mission that represents management incentive awards of $4 million, a noncash expense and $15 million elective cash acquisition of incremental equity stakes in several Mission numbers to buy back interest in the heads of those members. These management equity awards and equity buybacks are part of our strategy of long-term alignment with both the managers of the Mission business as well as a number of series through performance-based equity incentives. The Mission business continues to perform well beyond our expectations both the management incentivization and equity acquisitions represent significant long-term value for Accelerant. Over time, we expect to increase our ownership stakes in many of the well-performing Mission members beyond our initial investment of 70% to 80% ownership. On a segmental basis, we also outperformed our expectations across the board. Exchange Services is the core of our business, flowing from the 8% plus fee we make on all the premium running through our exchange. In Q4, we had revenue of $93 million, growing 46% year-over-year, well ahead of the premium growth over the same period because our take rate expanded to 8.4% from 7.2% in 2024. Exchange Services adjusted EBITDA was $63 million at a 67% margin. MGA operations is the host of MGAs we have ownership stakes in, consisting mostly of our Mission MGA incubation business. In Q4, we had revenue of $59 million with $2 million of that coming from the aforementioned nonrecurring investment gain. Excluding that gain, revenue was $57 million, growing 23% year-over-year. Adjusted EBITDA was $23 million or $21 million, excluding the gain, resulting in a margin of 36%. Our healthy operating cash flow and IPO proceeds have enabled us to maintain a strong balance sheet. At the end of the quarter, we held $524 million of cash in the entities outside of our Underwriting segment, including Exchange Services and MGA operations and $121 million of outstanding debt. As included in the appendix of our investor presentation, these segments generated 87% free cash flow conversion adjusting for IPO-related expenses and nonoperating cash flow investment gain. That added $157 million of unrestricted, unlevered free cash flow in 2025. Our business produces really high free cash flow conversion, even considering the investments we continue to make in the platform. We expect our free cash flow to grow nicely in 2026 as a result of growth across these 2 segments and the absence of IPO-related costs. Our Underwriting segment houses our own insurance and reinsurance companies. And as we have consistently stated, our goal is to limit the use of these entities and write more business directly with third-party insurers. The largest driver of Underwriting revenue is how much business we retain as net earned premium. In Q4, we had revenue of $111 million as our net retention of exchange written premium was 9% in the last 12 months. This resulted in adjusted EBITDA of $13 million. Our margin in the quarter was consistent with results in prior quarters, in part driven by good performance in the gross loss ratio at 51%. Our overall expectation of single-digit margin generation in the Underwriting segment has not changed in the medium term. We also expect in the near term that our net retention of exchange written premium will trend at similar levels, sub-10% that we have talked about previously. Within our Underwriting entities, we had $633 million of equity capital at the end of Q4. As we continue to accelerate the shift of premium to third-party insurers, we do not expect to need to put much, if any, additional capital into our Underwriting segment in 2026 and would expect to be in a surplus capital position over the long term. So in summary, Q4 was another step forward with strong growth and meaningful acceleration of adjusted EBITDA and adjusted net income, just like in Q3. As we look ahead, we have a lot of organic growth embedded in our pipeline. In Q1 2026, we expect exchange written premium of $1.07 billion to $1.13 billion, third-party direct written premium of $450 million to $470 million, and adjusted EBITDA of $64 million to $66 million. For the full year, '26, we expect exchange written premium of at least $5.1 billion and third-party direct written premium of at least $2.2 billion. To build a $2.2 billion of third-party direct written premium, we have $2.1 billion under contract and flowing today, another $100 million under contract that will start flowing in the coming months. We also expect adjusted EBITDA of at least $275 million, which does not assume any nonrecurring investment gain will recur during the year. It's important to highlight that we are expecting a greater proportion of third-party premium in the first quarter and full year 2026. As I mentioned before, this reduces the use of our risk-taking entities and increases EBITDA from our fee-based segments. Slide 13 of our investor presentation shows the expected EBITDA shift to more fee-based segments during 2026. Remember, when you think about '26 and beyond, our EBITDA growth engine is and will be the high-quality fee generation of the Risk Exchange accelerated and enabled by the AI-first strategy that is core to our DNA. We are not providing a GAAP reconciliation for this guidance, including net income due to the inherent difficulty in forecasting and quantifying items such as tax rate variations, foreign currency fluctuations or nonrecurring adjustments that may occur prior to quarter close. Lastly, our Board authorized a share repurchase program of up to $200 million effective through December of 2028. This authorization reflects our confidence in the strength of our balance sheet, the durability of our cash flows and the long-term value of our business. Consistent with our disciplined capital allocation, we believe share repurchases at today's trading levels are compelling and should be highly accretive as we continue to invest in our growth. With that, I'll turn things back to Jeff for closing remarks.

To take a step back and remind us all of the bigger picture, what matters in the long run is that Accelerant becomes the rails on which specialty insurance runs. We're making the best MGAs better with our data and analytics and connecting them to deep, high-quality risk capital. And we're doing more and more of that each year. The momentum is building in the market, the flywheel is accelerating. The future of specialty insurance is data-driven, increasingly autonomous and capital efficient. That future favors platforms built on proprietary data and architected for automation from the outset. We believe Accelerant sits squarely at that intersection. AI is a force multiplier for the model we've been building for years, and it's why I believe we're going to be the leading specialty insurance platform. 2025 was another exceptional year. We grew exchange premiums 35%, revenues 51% and adjusted EBITDA 149%. I want to thank our members, our risk capital partners, and our world-class team of experts that made it all possible. Before we move to your questions, let me welcome Cliff Jenks, our soon-to-be new General Counsel and Ray Iardella, our new Head of Investor Relations. Our team keeps getting better and better, and we couldn't be more excited to have them both on our side. Additionally, I would like to personally thank Jay Green for serving as Accelerant's Chief Financial Officer since 2022. He has been an invaluable partner, helping grow the business and successfully leading us through the IPO process. He recently informed the Board of his desire to step away from the business to pursue personal interest. On behalf of the Board and the entire team, thank you. We wish you all the best in the future, Jay. I'm excited to announce that Linda Huber has joined Accelerant and will be named CFO effective March 31. Linda is a very accomplished and seasoned executive. In fact, some of you likely know Linda from her previous CFO positions at leading business services firms, including FactSet, MSCI and Moody's. We are extremely pleased that a finance executive of Linda's caliber has joined the team and believe she will play a key role in our next chapter of growth. I know she is very much looking forward to engaging with the investment community in the future. I think we'll now open it up to questions.

[Operator Instructions] Your first question comes from the line of Rowland Mayor of RBC Capital Markets.

I just wanted to ask on the actual 4Q adjusted EBITDA. It came in almost 15% above the high end of the guidance range. I just wanted to understand kind of what the upside was in the quarter?

Great. Jay, do you want to take that please?

Sure. Yes, happy to. Yes, so I think upside, we really saw contribution from all 3 segments. So if you look at performance in Exchange Services, right? We beat on volume, right? And with the take rate going up, that contributed to the increase in EBITDA there. MGA Operations, again, we saw outperformance relative to expectations going in. And then, of course, greater contribution coming from Underwriting. We have -- even with lower net earned premium due to the retention being lower over time, the loss ratio coming in at 51% produced higher earnings there. So I wouldn't point to just one thing. I think we saw really strong performance across the board.

Okay. Perfect. And then I wanted to ask on the quarterly and annual guidance for the written premiums. I think it implies the second, third and fourth quarter have much higher growth than first quarter. Is that the MGA cancellation? And then can you remind me when that occurred and kind of what the size was in the premium through Canada?

That's right. It is higher. Ryan, do you want to give some of the details?

The size of the member generally, Rowland was around USD 50 million of premium per quarter. And so that is impacting Q1 and should also impact Q2 of this year. I think that's what baked into our guidance and our expectations. But you're exactly right on Q1 growth.

Your next question comes from the line of Elyse Greenspan of Wells Fargo.

My first question was on Hadron. I think I have in my notes that you guys had said it would be around 35% to 40% of third-party premium in '26. And I think going to below 33% in the fourth quarter and then drifting down over the longer term. Is that still the guidance when you laid out for everything for 2026 today?

Yes, we still feel very comfortable with those numbers. As you can see, we're making terrific progress there.

Okay. And then my second question, I guess, it's just on member count. Can you just give us a sense for '26, just how you're thinking about member count increasing throughout the year?

Ryan, do you want to take that?

Of course. Elyse at least generally, I would say the last, call it, 9 to 12 quarters, we've added roughly speaking, 15 new members a year. As you know, we try to sort of set expectations low and outperform. But our pipeline as we flagged, it's the largest we've ever seen. And so I think we're feeling good about our go-forward ability to add new members and grow.

Just one thing, Ryan. I think the context made it obvious. But if I heard it right, you said over the last 9 to 12 quarters, we've added 15 members per quarter.

That's correct. Yes. Sorry. Thank you, Jeff.

Your next question comes from the line of Mike Zaremski of BMO Capital Markets.

First question on cash flow. Thanks for the slide and the disclosure there. I heard in the prepared remarks about your commentary about not seeking to put a material amount of capital into the subs in '26, and you can correct me if that's incorrect. But my question is regarding how to think about cash flow on a forward basis. Do you guys expect kind of the trajectory to be similar to EBITDA growth? Or any context you can provide, I don't know if you think about it maybe as a percentage of revenues, et cetera.

Thanks. Before I ask Ryan to address those expectations, can I just highlight that it's an important turning point, at least in our mind, that we no longer will be putting material amounts of capital into those insurance company balance sheets, I think it should be noted. So forgive me, Mike, go ahead, Ryan.

Right. Thank you, [ Charlie ]. I think from a cash flow perspective, especially if you look at '24 versus '25, and I know we hadn't fully shared '24, but our cash flow conversion continues to improve. Obviously, we'd also expect not to have IPO-related costs next year -- this year, rather. And so I actually expect free cash flow conversion will continue to mix up and therefore, grow slightly faster than adjusted EBITDA. But we'll look forward to sharing those results in the future. The other thing I would point out, Charlie, and Jay mentioned this in the prepared remarks, but on Page 13 of the investor presentation, we pointed out, I think, how we expect 2026 adjusted EBITDA to come together. And what you'll see is, right, 45% year-over-year growth within the fee-based segments.

Got it. That's helpful. And definitely noted that we would have thought there would have been some capital being put at this up. So good news there. In terms of thinking about them the buyback program. Is that -- would there be potential not putting you guys -- not making you guys give us guidance, but would there be potential or the ability to upsize that based on free cash flow generation levels, that to the extent valuation remained attractive in terms of the share repurchase program?

Yes. So the question has 2 parts. We feel really good, as Ryan said, about the cash generation. And as that free cash gets generated, we and the board will have to make a decision about what's the best place to put that to work. And with the shares at anywhere even approximately near these levels, it's pretty hard to beat from our perspective and the Board's perspective. So it's early days, right, that we announced the buyback, obviously, last night. So it seems a little premature to be talking about what happens after that, but that would be our thinking.

Excellent. And lastly, the gross loss ratio remains excellent. Just from an industry-wide perspective, the entire industry continues to add a bit of reserves to other liability occurrence, especially. And so then my question is just kind of around the reserve addition, the small reserve addition, if you can give any context on the -- whether the gross and net impact were similar or just context about what took place there?

Yes. We feel really good about the gross loss ratios on each underwriting or accident year. Those numbers are terrific. The small adjustments between underwriting or accident years, I don't think is meaningful at all. Most importantly, to our risk capital, right? They're still receiving spectacular returns. Their enthusiasm to participate in the book continues to grow. So in all the ways that matter, we're doing great, as you said, on that loss ratio. I think the shifting around of reserves on the underwriting years, I don't think it's all that important.

Your next question comes from the line of Rob Cox, Goldman Sachs.

Thanks for all the helpful commentary on AI. And just a follow-up, given 95% of the accounts are sub-10,000 in premium. I just wanted you guys to maybe talk to us about what are the hurdles of this business going direct or through a more digital process than through a broker? And how can Accelerant members make sure they're still in line to win new business, if that happens?

Well, first of all, Rob, while we don't anticipate a wholesale shift of business, this commercial specialty business to go from a brokered transaction to a completely online transaction, whether that's AI-enabled or not. While we don't anticipate that to be the case. Were it to be the case, that would be great news for Accelerant and its members. The reason it would be great news is the production would be highly efficient, right? The data, which is the lifeblood of Accelerant's Risk Exchange would become that much cleaner and that much easier to process. I think given the fact that we are consistently the fastest response, so the first quote, I also think that given that our members consistently have the most specialized products, I think we and our members have a lot to gain if there were to be a shift away from retail brokers in this commercial space and towards some sort of automated AI-driven distribution.

Okay. Awesome. And then just on the new third-party insurer that was added since we all last spoke on a call like this. Can you just give us a little more color around what role you would expect that insurer to play over time in the context of the portfolio and maybe also like just the pipeline of third-party insurers?

Sure. Rob, one of the things that I'd point out is as we add third-party insurers, that are of really, really material size. We do our best to identify them clearly. So we've talked about Lloyd's, and we've talked about others in the past, as you know. This one that we've added, we're delighted to add them. We don't have their consent to use their name. They are going to participate on a couple of really interesting but specialty books of business that we'll see how big they grow. The important thing, though, is I think this story of adding this one that's highly specialized. Sort of talks about how we expect this develop -- to develop, I'm sorry. We have all the third-party insurer relationships we need to achieve our long-term guidance of 2/3 of the business going to third-party insurers. We don't need to make new relationships. Will we? of course, of course, we will. Some of them will be much more significant and much more across the board and some of them will be more specialized like the latest one this quarter. So that's sort of how we think about what the likely progression is. And then you asked about the pipeline. As the Risk Exchange continues to prove out its ability to produce this low volatility, very high profitability book of business. It's no surprise that interest levels increase. It's a -- the sales cycle is not short. I guess, is the best way to say it. And we will methodically add partners where we think that they're really additive and can improve the Risk Exchange. But I wouldn't want people to think that achieving our long-term goals requires us to create new relationships. It does not. We have sufficient relationships now. Ryan?

And just to be clear, Jeff kept saying long-term goals, just to be clear on the 2/3, that's a medium-term goal. I think you know that, Rob.

Yes.

I got the earnings call higher [ graphics ] wrong -- sorry, medium term. Medium term.

Your next question comes from the line of Paul Newsome of Piper Sandler.

I was hoping if you could give us a little bit more commentary around the expansion in the captive businesses? And just to sort of maybe give us a better sense of how the risks and trade-offs of that business compare with sort of a traditional MGA kind of source premium stuff?

Sure. We're pleased with the development, the development we've seen on the captive book. I think we've said in the past that we expect in 2026, the captive book to contribute over $100 million of premium. That's growing to what we think over the medium term to be a really significant contributor to the overall portfolio. Now in the context of your question is the fact that if you did a market analysis of where does most captive premium dollars come from, they come from very, very large companies retaining very large portions of their significant complicated, sophisticated risks. That's not where we try to play. Where we try to play is in the again, small to medium-sized companies trying to be more efficient in how they buy their insurance. And keep in mind, the real driver here is when the insured or the sponsor of the captive feels like the market insurance rate is mispricing them, obviously too high. So we get really excited about opportunities to go after self-selected, outperforming books of business. I would say that the lines of business are the same as our overall portfolio. I would say the limits issued are very, very similar. We avoid the same things that we avoid in our traditional MGA book being coastal catastrophe exposures than anything that aggregates with the industry portfolio at large. So I would say, while not perfectly homogeneous, they're really, really close. And the key there is we're avoiding the big giant companies captives.

Your next question comes from the line of Craig Maurer of FT Partners.

Most of my questions have been asked and answered. But I wanted to ask about the net retention ratio. It's bounced around a little bit in recent quarters, but below the 10% target. I was wondering if there's a natural limit to how low that can go as you continue to shift to fee-based segments?

As we described, the vast majority of our net retention is now driven by different jurisdictions, regulatory minimums. So obviously, there's not much we can do about those. As we -- what we've said is we'll be around but slightly below 10%. And it does jump and there's nothing we can do about that because of mix. But you're quite right that as the business more and more goes to third-party insurers, what I think -- what I anticipate saying someday is, it's going to jump around 5% as the overall share of exchange written premium goes down for the Accelerant-owned carriers. Does that make sense?

Yes.

Our next question comes from the line of -- [Operator Instructions] We don't have any pending questions. I'd now like to hand the call back to Mr. Jeff Radke for final remarks.

Great. Thank you very much, and thank you all for being with us this morning. We have significant momentum, a world-class team and are executing our strategy to grow the Accelerant Risk Exchange. We look forward to updating you on our progress throughout the year. Thank you again. Thanks, operator.

Thank you for attending today's call. You may now disconnect. Goodbye.