All right. Good morning, everybody. We'll keep this on time. Thank you for being here. Day two of our Global Financial Services Conference. For those of you who don't know me, my name is Manav Patnaik. I cover business and information services for Barclays. One of the companies under our coverage is Verisk, and we're very pleased to have Elizabeth Mann, the CFO, back with us here again. Thank you for being here, Elizabeth.
Thanks so much, Manav, for having us, and thank you all for coming.
I thought the best way to start would be to just talk about the current trends implied in your guide, the long-term guide. Maybe before that, let's just knock the question around the AccuLynx deal out of the way. Just the kind of latest update with the situation there.
Yeah. I think we've announced the AccuLynx transaction, which had been under litigation. We have appealed the decision now. That is under appeal. The hearings would start in late October. It is continuing to be handled on an expedited basis as the original trial was.
Got it. Okay. Fair enough. Maybe just getting back to this year's guide, maybe just remind us how the first half performance was, and then kind of the implied second half guide in there.
Yeah. Thanks a bunch. Yeah. Our revenue for the year, the first half, had been impacted by some near-term revenue headwinds that we have talked about. I am sure we will get into that. Short-term factors, the weather impact being the most significant, and primarily on the transactional revenue growth side, which is 17% of our revenues. Our subscription growth, at 83% of our revenues, continues to perform well. Then to your question on the second half, we have continued to say our expectation is that those short-term factors, we are working through them and working past the time that they would have an impact.
We expect to see continued accelerating growth into the second half of the year, with an expectation that we deliver within our quote-unquote, "our long term or Investor Day guidance range of 6%-8% organic constant currency growth." We expect to be in that range in the second half of the year. We have also said we are confident to deliver another year in that range, meaning the full year 2026 to be in that range.
Got it. Maybe just first follow up on the subscription side of the business. I mean, the first half of the year, the numbers actually got better, despite some of what we thought were probably tough comps heading in. What are some of the drivers of that slight acceleration in that subscription growth?
Yeah, I think we continue to see good outcomes and good interest from our customers on the value that's being delivered across the platform. It's not just one business or one upgrade. It is really across our largest businesses. Our forms, rules, and loss cost business, which is our largest business, continues to see the benefit of Core Lines Reimagined and the significant technology investments that we've put into our core products, including AI-based innovations. So that continues to express itself in the subscription growth there. Our catastrophe and risk solutions business just saw the launch of Verisk Synergy Studio, the new platform there, as well as the new model for U.S. hurricane risk. So that's seeing strong subscription growth as well.
Got it. I guess on both those, it sounds like you have a new platform and then Core Lines Reimagined is also new capabilities. Is there scope for further acceleration driven by those two? I guess, how early in the innings are we on those two contributors?
Yeah. I think we will continue to see the benefit of those play out over a long term, meaning a couple of years. Both of those businesses are some of our largest businesses, have very long-term contracts with customers, meaning three years to five years for some of the largest customers. So the benefits and enhancements that we've built in are being monetized in the subscription renewals. So we do anticipate continued strength.
Okay. But it's also fair to say that the current elevated or higher growth rate at subscription is sustainable. Then as the contracts come up for renewal, I suppose you can keep adding on more of the value in there. Is that fair?
That would absolutely be the intention. We don't give specific-
Yeah
-specific guidance by subscription or transactional revenue. We're targeting the total. But yes, absolutely, we do continue to see positive engagement with our customers.
Got it. Then maybe just touching on Core Lines Reimagined, it was obviously a heavy lift for four to five years for you guys. I know it never ends, but what is the end outcome today, where we stand, and what's the runway of the potential for going through this effort?
Yeah, absolutely. Core Lines Reimagined was reimagining the benefit and really primarily the delivery mechanisms and the underlying infrastructure for our forms, rules, and loss cost business, really reinvesting in a product that had been deeply embedded in the insurance industry, but maybe hadn't gotten an overhaul in quite some time. That was a five-year scope. We started it in 2021. By the end of this year, we will have delivered the full program that was envisioned then. Most of the new and newly rebuilt content is available on core.verisk.com. We continue to turn things on module by module and line of business by line of business. By the end of 2026, it will all be available on core.verisk.com, and you won't have a need to go back to the old ISOnet platform.
There may be still some customers who prefer to use it that way, but we will be pushing hard to shift them over to core.verisk.com. That will be kind of the We will draw a line under the completion of the original scope of Core Lines Reimagined. It has been such a benefit and such a success. When that was designed in 2021, we were doing a lot of machine learning. We were doing a lot of AI. But it wasn't an environment where everybody was talking about Generative AI. We've brought some of those elements into our platform already. But I think where we stand now, we will continue to see benefit. We're not going to stop investing in the product. The investment scale may taper a bit.
But we continue to see significant opportunity and significant client interest in what can be done off that core content that is so valuable to them.
Got it. And maybe just a few examples on that, because part of, I think, what you described felt almost like a necessary modernization of the way you had the data, the delivery, and the platform. Now that you have that, can you talk about whether that helps innovation stuff, companies? Give me just some examples there.
Yeah. Let me give some examples. The product there it is the loss cost, meaning the industry average-wide expected loss on a particular type of risk with deep granularity, not just by class of insurance general liability, but by business type, say small commercial, by state, and by type of risk. So it's not just commercial liability for a restaurant. It is separated out into 13 different grades of restaurant by state. So very granular views of risk based on the industry-wide average and/or industry-wide views of the risk. And really no one carrier has enough penetration or sees enough data. That's the core value of the product. The delivery method was where, as we talked about, it had maybe been under-invested.
And so if you go back before Core Lines Reimagined, those loss costs were primarily delivered by PDF in tables that customers would download and probably manually key in into their own internal systems or into their own workflow. That we knew was not how work should take place in the 2020s, even before anyone started talking about AI interaction with it. We rebuilt the platform, we rebuilt the underlying data. I should say this was all enabled because in the 2010s we moved that core data set to the cloud and on modern database infrastructure. So that was almost step zero of Core Lines Reimagined in the 2017 to 2021 era. So we've got our database our industry-wide data housed and manipulated in modern data methods.
Of course, this was also an important prerequisite in being able to use and handle that data with AI and other modern methods. So delivery mechanism is more modern. Customers only download the PDF if that's still how they want to consume it. They can now access the data via API. They can access it through accelerators and integrators into their policy administration systems like a Guidewire or a Duck Creek . Today, if they choose to, they can interact with that data via an MCP using Quad.
Okay.
So those are some of the different delivery mechanisms. The forms content, the policy language that is so crucial for carriers. Insurance, each policy is a legally binding contract with language that really matters. It matters first because it has to be approved by the regulator on every line of business in every state that you're doing business in annually. Your policy form has to be approved by the regulator. Updates and changes to it also need to be approved, and then it matters also because that language is court tested. It sets policies. It governs the liability for what carriers are exposed to on what risks. So this is very legally sensitive language. It remains legally sensitive.
What hasn't changed on our rules business is the number of actuarial experts, legal experts, local legislative experts that we have that is giving input into that forms language, which is far more efficient for us to do once versus each carrier in the industry doing that analysis and hiring those people themselves. That hasn't changed. What has changed now is the ability to interact with the forms language in much more modern ways. Again, not just working with a PDF. We actually built a data structure underlying those forms and that policy language a couple of years ago, again, before ChatGPT and other things. So we have a modular data form underlying the forms and the policy language, which enables much easier and better interaction with that form. And so you were asking for specific examples of how does this save time.
It used to be even five years ago if an underwriter at a carrier was taking a look at this year's policy, they said, "Okay, I read there was a legislative change in Illinois. It's going to change my commercial property form language." They would download this year's form. They would have to search through and find the relevant paragraph and figure out how the language had changed. Now it is much more digitally interactable. You've got a map. You've got legislative monitoring. You can click on a state, what legislative changes were made, click through what policies have changed. Click through actually have these policies changed in other states and where has it changed my exposure? So those are all the things that are enabled now.
In the forms business that have been enabled by Core Lines Reimagined, and you can easily imagine and extrapolate from that how carriers might use GenAI to interact with that to build their own modifications on top of our forms, or to create their innovations on forms language. To be clear, any innovations that they make with those forms, number one, will need to be also filed and approved by the regulator. They can do it themselves from scratch, but that would mean taking legal liability on the full form. It would mean having that infrastructure to build the full form, and it would mean full regulatory review of that new policy. Or they can start with the industry-wide gold standard starting point. They can download that.
They can say, "I've tweaked it in these three places." Because of our infrastructure with Core Lines Reimagined, it's much easier to identify where those three places are that they may want to tweak, and they can file that for approval and get reviewed just on those three modifications as opposed to the full form from scratch.
Got it. Okay. Thank you for that.
Sorry for that long answer.
No, that is fine.
I wanted to bring to light what is one of our really core businesses and our core value proposition and what we continue to provide to the industry that is still valuable even in this world.
Got it. One of the other initiatives you mentioned briefly there was the kind of MCP delivery of data. Can you just help us appreciate how much of the data that you guys have today is available through that and what the strategy with MCP is going to be going forward?
The MCP that we have today is for convenience on the part of our carriers. What they see is the loss cost, which is the table, the industry-wide average. That is what we would call the output. Maybe you would call that an analytic. It is based on the internal raw data that is contributed across the industry. They can see, they can interact with that loss cost data. It is the same thing that they get access to in their core subscription, but now they can ask questions of it, why were loss costs trending down in Florida over the last couple of years? It is because of regulatory reform. They can ask business-related questions. They can integrate it into some of their other things. But I think the key thing to know is that is the loss cost table.
That is our IP that is created and being monetized and accessed through the MCP connector. It is not the raw data that is contributed across the industry.
Got it. So when you say for convenience, I guess it is just another delivery mechanism.
That is right.
But there's no uplift in pricing or some kind of a premium attributed to accessing that data?
At the moment, no. We are assessing the value that is created for the industry in accessing the data in that way and interacting with the data in that way. If we believe there is value that's being created as they use it and interact with it, we will charge for it, and we will monetize that value.
Okay. Then maybe a couple of other examples of AI. I think on the call you mentioned XactAI. So what is the uplift or what is the new features of that?
Yeah. So the new features on XactAI is a number of AI tools that is added to the core software, XactAnalysis or Xactimate. That's the software in our Property and Restoration Solutions business. It's used by the carriers themselves and the contractors or third-party adjusters to develop an estimate and approve an estimate for property repair associated with an insurance claim. So a tree falls on a house, maybe in a hurricane, maybe not. The contractor goes on site, the insurance adjuster goes on site, and they build an estimate. They say, "Okay, here, let me assess the damage. Let me take photographs of the damage." Then I build an estimate, and it's going to take, it was three windows that were broken. I need this much wood for the frames. I need this many panes of glass.
The labor, which is a very hard part to develop pricing on. This is a job that's going to take three people you know six hours of work each. Here's the cost of that in this region at this moment in time. It is both the software that develops that estimate. It is the pricing data, both materials and labor, that populates that estimate. That's what happens. The contractor, the adjuster, develop the estimate. It goes to the carrier. The carrier approves it. The carrier assesses, is it in line with my rules and policies and compliance? Is this an appropriate estimate in this region and for this type of work? The AI features enable the process and the development and those processing of that estimate much more quickly. Now if you take a photo, it can auto-populate. Oh, I see, it's this type of window.
It was this type of glass, and it will suggest, this is what you're going to need, and here's the price from the pricing estimate. Obviously, that gets reviewed and adjusted. Compare that to, it saves a significant amount of time for the adjuster versus taking a picture, going back, having to look at a price list, having to look at trying to assess what each element in that photo is. Maybe forgetting a couple things and having to go back to it later.
Got it. Maybe just one moment to go back to the Verisk Synergy Studio. One of your main competitors has also introduced or has introduced their own platform.
Yeah.
Is there a trend in the industry? Is it just AI capabilities allowing for this platform or what? Then also just for you, how should we think about whether that starts creating uplift in that specific business?
Yeah. The platform, our platform or our competitor's platform, were not designed as AI-specific innovations, but they already do and may have further AI enhancements, or obviously AI was used to develop the platform. I think it was more that they were built and created to have more modern and much more powerful infrastructure on which to run the catastrophe models. It used to be the catastrophe models were run on on-prem, accessed on a platform that was an on-prem build. Now it is truly SaaS and can use as much power or compute as you might want. I think industry standard used to be to run 10,000 paths on a Monte Carlo. But if you want it, if it's appropriate to have a more robust view, you can now run 100,000 views of risk.
You can take and modify each of the parameters associated with the risk or with the portfolio that you have. It's much more interactable and you can kind of take your own view on risk, and we can say, "Well, look, our expectation for the assumptions that you should use are this," but you can kind of take and modify those platforms.
Okay. The rollout of the studio, I know your competitors talked about it'll be a nice pricing uplift for many years. Is it the same for you guys or?
Yeah.
Okay. Maybe just putting all this innovation, I'm sure there's more, but together, it sounds like a lot of it is being monetized through kind of your pricing uplifts in your renewals as opposed to directly incremental revenue. Is that fair, or how would you kind of break that out?
The ones that we've talked about, Core Lines Reimagined, Catastrophe Risk Solutions, those are fairly mature products that are deeply embedded in the industry, so there's an expectation of what the product is. Yes, those innovations will be monetized through pricing on the core subscription. Something like an XactAI, which is more of an add-on, that is a separate SKU, a separate revenue line that a customer can opt into or opt out of, and pay for accordingly. So that you'll start to see the monetization-
Okay
-over time.
Then maybe somewhat tied to this, if you go back to the last Investor Day, you actually raised the component of your growth that was pricing by 50 basis points.
Yeah.
Can you just talk about the reason behind that? Maybe it was anticipation of all this innovation. Just curious.
Exactly. It is both anticipation of the innovation as well as increased comfort that we have in being able to monetize that based on the demonstrated success of what we had done with primarily Core Lines Reimagined before that. If I rewind, in 2023, we had said we expected pricing to be 3 percentage points- 4 percentage points of our 6%- 8% growth. We actually delivered right about 5%, and that was both monetizing product innovation through pricing, yes. It was also, in fairness, had a contribution from the strong premium growth environment, which I am sure we will get to.
Yes.
That helped the price increases there. With that confidence of having significantly over-delivered on that range, we raised the range from 3.5%- 4.5%. If you like, 3.5 percentage points of annual growth we would target to come from price. If you like, that gives room for a little bit less pricing benefit from the premium environment versus the 5% we had already delivered, while saying that we believe the pricing benefits that we get from product innovation, we expect to be sustainable and to continue.
Got it. That 5%, the 100 basis points- 200 basis point outperformance versus your 2023 guide, can you attribute how much of that was premiums versus your own initiatives?
We have not quantified it, and it is actually not concrete or formulaic from a premium standpoint. I could not tell you even if I wanted, but it was clearly both were in there and that was our framework for it.
Okay. And so maybe let's just touch on the premium topic, and maybe rehash again as we do every year what the relation to the net written premium number is versus your contracts and how you look through those.
Yeah. So we set the stage, when we became independent from the industry, that we would grow as our customers grow. Our contract renewals had an input into the renewal price increase that came from the customer's premium growth. I think that's a great revenue model. It ties our growth to an externally observable growing market, which is the insurance and the insurance premium market. While any given year it may vary a bit, and we've had a couple years of very strong premium growth. But on average, we've got an input into our pricing that is growing low to mid-single digits. So that is a contributor to our pricing growth. We also have the opportunity to take our own Verisk price increases in addition to the premium growth. So both of those things are inputs into the growth rate.
We say it's about 20%-25% of our revenues now that are on contracts that have some input from that premium growth component. It's not directly formulaic, but it is a supporter of that. So we probably grow faster in stronger premium environments, and a bit more moderated in softer premium environments. The one other thing I should add is there's a two year lag on that premium growth dynamic.
Got it. And maybe if you could just remind us of the stat you'd given before in terms of, you said in the softer market, which is what we're seeing now, you would expect a moderation of growth. How much is moderation implied?
Yeah. The historical view, and each year is different and there's always different dynamics. But when we went back, we said this at our Investor Day, when we went back since we went public in 2009, and we classified the years into soft market years or hard market years. The years that were soft market years, our average, and this is just organic constant currency growth in our insurance business throughout this time, it was 6.8% in the soft market years. It was 7.3% in the hard market years. You can see in there's a bit of a headwind when the market is softer or a bit of an acceleration in a hard market. But at the end of the day, it's only moderate, and both of those average out to our 7% historical average growth rate.
Got it. I do not remember the premium growth in those previews that you talked about, but for example, you said the last few years have been really good premiums, right? In the upper high single digits, and now you said they're low to mid. I think the first half was low single digits thus far. Is that delta also encapsulated within that?
That's about the range. I do not know the exact averages in those years, but the highest years were high single digits. We may have maybe one year tipped, just barely touched double-d igits. That's the high end of the range. The low end of the range, at some years in that time period, you've actually seen premium growth being negative, and we've continued to grow even in those environments.
Got it.
But more recently, the last couple of years have been high single digit premium growth, which is very high, and you have probably read and heard about this, if not experienced it yourself as an insurance consumer. This year it is moderating. So for the first half of 2026, we are seeing low single digit premium growth. It varies line by line of business as well.
Got it. Maybe we will switch quickly to the topic of weather. I think in the last update you said weather was tracking below average or light versus what you had assumed, I suppose. So maybe just some more color, flush out exactly what those assumptions are and what you are seeing today.
Yeah. So our Property Restoration Solutions business that I talked about before gets revenue by helping process claims associated with property repair. And so it can get benefits in very active weather years, which certainly was the case in 2023 and 2024. 2025, as it happened, was a very light weather year. It was the first year in a decade that no named hurricanes hit landfall in the U.S. That is a good thing for homeowners and a good thing for policyholders and a good thing for insurance carriers. It was a headwind for revenue associated with claims. I want to be clear, our business is mostly subscription, and even in that business, we have been converting more and more customers to a subscription basis. So the transactional piece is kind of the incremental piece.
But because of the stability of our revenues, people have been watching that transactional revenue as it is maybe one of the swing factors. But with a business as stable as ours, 25 basis points or 50 basis points moves the needle. So we have been watching those weather activities. Because of the subscription nature of our business and because this is only incremental, we plan and forecast for, quote, "an average year of weather," which is what we have done consistently throughout.
In 2023 and 2024, we significantly outperformed that average year of weather. 2025 was a much lighter weather year. Going into 2026, I would say we planned for an average year of weather. So far, it appears to be a bit on the lighter side. That is not necessarily a headwind versus 2025. To be clear, we do not anticipate that it would It did not change our guidance range.
This is just modest variability within the range. I think the trends that we're seeing right now in mid-September, is playing out as roughly in line with what we saw and what we expected as of the end of July when we spoke at the end of our second quarter call. So relatively light. It's mid-September. There hasn't yet been a major hurricane in the U.S., which again, is a good thing and is in line with how we talked about our guidance before.
Got it. Okay. Thank you for that. Maybe this is a good time just to help us appreciate what all is in the transaction business, because I know subscription's the majority, but you get, I think, the majority of the attention on the transaction side.
Yeah.
What is the mix of the different items in the transaction business today?
Yeah. Let me start with our goal is to maximize revenue across the portfolio, and our goal for each of our products and businesses is for them to develop and establish the revenue model that works best for that business and its customers. Each of our businesses tends to evolve to a mix of subscription and transactional. Across the portfolio, again, we're about 83% subscription, 17% transactional. What's in that transactional revenue? There are some businesses where it's really more of a usage-based or a volume-based business, and we have a traditional price volume. That can be some of our underwriting data products. So in the UDAS part of the pie can have products where it is a data pull that is typically correlated with an underwriter underwriting a piece of business.
On the claims side, we have some businesses that are, as I said, associated with volumes of typically number of claims. That is often a subscription with an overage tier that correlates with the number of claims. The other area that we have is services businesses, so where we have some software businesses on the life insurance side, on the Specialty Business Solutions side, we have a services component. These are not heavy implementations. They are pretty quick, but we have found it enables our customers easier access, and improves the sale of the license revenue to have some support in implementation. So we have got some services revenues in there. Then the final piece, which falls in that volume category, is the weather elements claims associated with the weather. Those are the largest parts of our transactional revenue portfolio.
Okay. That is helpful. Maybe just on the broader topic of GenAI, I think at your Investor Day, you did a pretty good job in outlining how much of it is proprietary data, et cetera. Just a few lingering questions that we get. So the first one tied to you guys is more, over the years, Verisk has talked about the softwarization of the business, so the question is more how much of that is just true software that could be vibe coded away versus how much of that is integrated with your proprietary data?
Yeah. I think, we gave the pie chart at the Investor Day. A lot of our products are built on contributory data sets from across the industry, which we view as very differentiated. There is another significant part that is proprietary data, where we are the only ones that, or where we create it and have it in-house. It is not easily vibe coding away. There may be, in some cases, a couple other providers of similar data, but we have strong performance, and strong proprietary data there. We had another category that we called proprietary analytics and IP, and those are things that are not exactly data businesses, but are built on very deep domain knowledge that it would be very difficult to replicate. The catastrophe models, for example, would be in that category. Or we talked about the Xactware, the Property Restoration Solutions business.
Some of that business, yes, is software, or it is delivered as software, but it is software that is built on top of and enables the access to the pricing database, which is proprietary and very difficult to replicate, certainly, for example, on the labor pricing side. So that would go in that bucket. Those categories together amount to 85% of our revenues. There is only 10% of our business that we would call true software and another maybe 5% that are services-related businesses in the casualty space or in the international space. So those would be the ones. The software, in theory, you could say, oh, maybe that 10% could be vibe coding away. We still think it is either based on deep domain knowledge in very regulatorily sensitive spaces, like life insurance, for example, or it is software that is protected by a network characteristic.
For example, in our Specialty Business Solutions product, where the software, again, is enabling something that is transactions between a broker and a carrier, and the software is where they come together to transact. We may have businesses that are software, but what's differentiated may not be the software itself. It's more the characteristics of the business, where it's a network type business, and it's where the industry comes to interact, to exchange financial transactions.
Got it. The other one we get is, like at this conference, I think a lot of large carriers are here, and I guess they tell investors, "Yeah, we're going to use AI to do a lot more with our data and bring stuff in-house." In your discussions, does that impact your business? Does it create opportunities? Just how would you react to that?
Yeah. We think it creates opportunities for us. For many decades, Verisk has been a strong participant in the industry. We contribute a lot of data. We contribute a lot of insights and value to them. From a scale perspective, our revenues are 30 basis points of the total industry premium. So for every $100 a carrier gets in premium, they spend $0.30 on the Verisk data. We think that is good value to them, to protect the insights and the views on selecting the right risk, pricing it appropriately, and then processing the claims accordingly. As you can see, there's many things outside of that $0.30 that they may do on their own. They may each take different approaches. The large carriers will have more bandwidth to invest and will do some things with their own data.
The small and midsize carriers may be interested in doing some of the same things, but might be interested in productizing different products from us to enable them to have the same reach and scale.
Sure.
I think another thing that may be underappreciated in this is we are already speaking across the industry. We have that contributory data angle that gives us a perspective on the full industry that we can share back to them and create value, even for those that are doing things in-house. When you think about the fragmentation of the industry, insurance is a very large market. The largest players in their largest lines may have 20% or max 30% of the market in a certain line. I have been in rooms with customers that say, "That's great. I have 70% of the market that I don't see, and I need to know what's happening there to be able to price the best risk." That is our fundamental value proposition that we think will be maintained as an element of even what the large carriers are doing.
Got it. In the two minutes we have left, maybe let's end with capital allocation. Since you backed out of the AccuLynx deal, you did a big bunch of buybacks. Maybe just remind us of current leverage levels and the plans for buybacks and other capital allocation priorities.
Yeah. Leverage at the end of the second quarter was 2.5x debt to EBITDA. That is at the midpoint, essentially, of our 2x- 3x debt to EBITDA range. We have significant and growing free cash flow. We are fortunate to have significant capital allocation firepower, capacity, and bandwidth. We have the opportunity to continue buying back shares in the market. Yes, the AccuLynx appeal is pending, and we'll have to see how that comes out from a large allocation standpoint, but we're very comfortable with our leverage range. If, and this is a big if, but if that transaction were to ultimately go on and be completed, we have the bandwidth to do so on an all-cash basis, and get back to our 2x- 3x leverage range within a relatively short period of time.
Okay. All right. We'll end it right there. Thank you so much, Elizabeth. Appreciate the time.
Thank you.
Thanks, everybody.
Thanks very much for attending.