Verisk Analytics, Inc. ($VRSK)

All right. Thank you, everyone, for joining. I'm Andrew Nicholas. I'm the business services analyst here at William Blair. Before getting started, I am required to inform you that for a complete list of research disclosures or potential conflicts of interest, please visit our website at williamblair.com. With that out of the way, I'm very pleased to welcome Verisk's CEO, Lee Shavel, to the 46th Annual William Blair Growth Stock Conference. Thanks, Lee, for being here. I'm going to hand it over to Lee to give you an overview of the business.

Good morning, everyone. Thanks for taking the time. I'm sorry, I'm getting adjusted to the microphone here. But thanks for taking the time to meet with us today. I'm going to make sure the slides are working here. As an overview for Verisk, I want to try to accomplish three things today. The first is to give you a clear sense of the underlying core durable franchise that Verisk represents. Two, to help you understand the growth opportunity that we have to pursue to sustain the growth that we've been able to deliver with some very distinct competitive advantages and then tying that to what we think is a very strongly differentiated financial model, which has consistently delivered value for our clients over time. And for our audio specialists, is this mic upfront on? I want to make sure it's not interacting with -- okay. Excellent. Thank you. So first, for those of you that are not familiar with the Verisk model, we have started as a utility to the U.S. property and casualty insurance industry over 30 years ago. We are entirely focused on insurance. We have about $3 billion in revenue and the distinguishing characteristics of this is a highly recurring revenue model that is largely subscription-based. And that forms a very solid foundation for the economic model of what we -- of how we provide value to the industry. And what underlies the strength of that business model is the fact that Verisk sits at the center of the global insurance industry. We serve all mentions of both risks, as you can see on the top side from a property, auto and both from a functional underwriting, claims, and even in related insurance areas such as property repair, estimating and fraud detection. And so these are all functions across the industry that we serve. And beneath all of these functions is really the core asset of Verisk, which is the connective data that Verisk gathers, normalizes and integrates into the dozens of products and functions that we use to support the overall insurance industry. And kind of building on that prior slide, you can see the individual products that support each of these distinct areas that we are able -- where we are able to support a major global, national, regional, local market insurance provider across a variety of their products and functions. And I think what is most important and certainly the question that were asked most frequently is what is the durability and sustainability of those data sets. And as we discussed at our Investor Day, over 90% of Verisk's revenue is underpinned by a unique data advantage. Where, for instance, 40% of our revenues are tied to data that is directly contributed to us by our clients. Now this data isn't usable as they give it to us, it requires a substantial amount of actuarial normalization because each carrier underwrites to different products to different standards, with different amendments, different aspects of their policies that we need to normalize, and this isn't a theoretical exercise, it also is one that has to meet regulatory standards because the data sets that we are providing to rate policies are used as the primary basis for rate approvals by regulators. And so it has to be actuarially sound. It has to be scientifically rigorous and this data that we're collecting can't be acquired anywhere else. It's not publicly available. It's highly sensitive data that the carriers trust us with, and it requires a massive amount of work to normalize this data for use in these functions. Now in addition, we also gather our own data sets. So for instance, in this building and pretty much and any other building that you can see from the hotel here, Verisk has a field team that gathers detailed engineering data, fire suppression data, ground floor use and function data for this building that is very relevant to insurers that are ensuring either the building or businesses within the building. But not only that data that we gather on a proprietary basis, we also rate every fire department within the U.S. So the Chicago fire department receives a rating from Verisk on the basis of the quality of their equipment, the quality of their training, their speed of response. Not only do we provide that, we provide a building code effectiveness score. So in every municipality, it's not enough to understand what is the building code, but is it actually being enforced? And is it effective? Critical to understanding, particularly how catastrophes can potentially generate risk. So these are -- this is an example of proprietary data that we collect with expertise specific to the insurance industry and importantly, a competitive advantage is that we have an ability to gather this data at a much lower cost than the industry would incur if they were gathering this individually. So there's a natural economic efficiency that exists. And then the third primary component is our proprietary intellectual property and analytics. An example of this is the scientific intellectual property that we generate in modeling, tropical cyclones or hurricanes, severe convective storms, wildfires, floods, and catastrophe modeling is not weather forecasting, whether forecasting is a component, but it's understanding how weather and insured assets interact. And there's a high degree of structural knowledge, materials, labor costs, exposures, risk mitigation that goes into that. And our models serve as a standard for reinsurers and carriers and brokers to facilitate the transfer of risk within those markets. And that expertise is not easily obtained or integrated or certainly to set a standard that the market relies upon on a daily basis. So I spent a little bit more time on these data components because they're really important in this market environment where there's a lot of anxiety about AI disruption, and we certainly recognize that risk. But these are not public data sets. They're utilized for both regulatory and industry standard support and are very difficult to disrupt. Now that leads us to a high degree of resilience and predictability as you can see on this review of our revenue growth. And you can see across this period of 15 years that we have seen a steady growth in our cumulative annual growth rate. And it's a function of both the increasing adoption of the use of data and analytics to automate and improve the efficiency of the insurance industry. So if you think about the percentage of the use of data relative to traditional human means of evaluating risk, data, technology and analytics has become more important. It also has reflected the adoption of new technologies that facilitate a variety of industry functions, such as claims anti-fraud capabilities or the integration and facilitation of specialty risk transfer in the Lloyd's market, where we've developed software platforms and ecosystems to facilitate that. So we continue to see the industry adopting more technology, and we're in a unique position with our centrality and our scale to deliver on that. All of that growth, if we look at it on a purely organic basis, has generated a very consistent 6% to 8% organic growth rate, where we have been in that range and only been below it in two instances in 2009 in the midst of the global financial crisis, where we still generated a 5% organic growth rate. And in the initial year of the disruption of the pandemic, where, again, we were still generating a 5% organic growth rate. And certainly, when you look at these numbers, you can't avoid thinking about the compound power of delivering that consistent growth over time and particularly enhanced with the operating leverage that we'll talk about when we give the financial model. So with that foundation for our core, let's turn to the growth opportunity that we have ahead of us. Now a foundation, but by no means a governor for our growth opportunity is the overall growth in the property and casualty insurance industry. And you can see here that over the past 15 years or so, the industry's gross U.S. net written premium has been growing at about a 4% compound annual growth rate and has -- is projected on the basis of projections by the NAIC, the National Association of Insurance Commissioners as well as Swiss Re is expected to increase to approximately a 5% growth rate over the next 5 years. Now there are a couple of factors that are driving that growth. One is just more greater -- more risks or more assets that need to be covered and naturally inflationary factors that grow that. But within that growth, we benefit from specific market forces that are driving more demand for data. One, as I mentioned before, carriers are looking to automate more and more workflows through automation with data sets. And AI specifically is driving the demand for greater data. AI without high-quality data sets cannot be nearly as effective. And while the focus has been in selling broad enterprise licenses for large language models, what's critical is tying that to existing workflows and data sets. This is not only our opinion, but the opinion of both our clients as well as the frontier model companies that have expressed a desire to worth us much more actively realizing that the returns from improving the efficiency of an underwriter or a claim is going to be critical to demonstrating a real return on the AI investment and certainly was a foundation for our opportunity to develop two connectors with Anthropic for their cloud model to just two of our data sets. So the carriers are certainly driving more utilization of data. We also see in the market that the use of data for intermediaries like brokers or MGAs or even regulators is important. We're also seeing expansion in markets such as the excess and surplus market, which is an adjacent form of insurance that supplements the regulated or admitted markets where clients are giving us new data sets. And then finally, more macro trends such as affordability and climate issues and how they affect insurance markets are creating demand for new data sets. Now with both the growth in the underlying market and these trends Verisk brings several significant advantages in our ability to capitalize on that on each of these. And I'll talk about each of these in turn. The first is that we are a mission-critical and a trusted partner for the global insurance industry. Our data is the backbone of the underwriting function, particularly related to regulatory rate filings. Our data is used as an actuarially reliable source of their basis for rate increases and rate adjustments. We have broader data sets than any other company normalized for their specific purposes. And we've been playing that role for decades. And I can't understate the importance of the trust in protecting the data and utilizing the data for the industry's benefit that is a substantial advantage and a natural starting point for us as a partner. And that differentiation, I think, is emphasized on this slide. It's a busy slide, but what I'd like to emphasize is at the center, Verisk provides as an innovation hub, the scale, the quality and the breadth of data and our ability to experiment with new technologies in a way that the industry can utilize to get comfortable and to generate higher returns on their investment in technology because we have these broad data sets that we source from across the insurance ecosystem and we're able to translate them into very tangible benefits in underwriting claims, catastrophe modeling compliance as you can see, improving 80% faster growth, 3% to 4% pricing uplift, more accurate estimates, faster settlement, all of which translate into value for the industry. And our opportunity is to make that investment, test it, prove it and deliver it to the industry. And certainly, the game plan that we have been pursuing with AI across several fronts, as we'll talk about in more detail. Our strategy has been one to engage in AI, so that we can learn from deploying it internally against our data sets to enhance our own analytics and generating -- and enhancing our insights of what we see across the industry. We have integrated into over 60 of our products that have allowed us to test that with clients. You can see examples that we've talked about, each of these, you can review on our website. And most recently, it's enabled us to put together in a very short order, a partnership with Anthropic, where we have been able to connect their cloud model to interrogate our analytics and facilitate and insurance underwriting professionals, questioning of what's happening from a regulatory and from a loss cost standpoint as well as to evaluate and determine what the potential restoration costs are on a remodel or repair of an existing business. Now that certainly was an exciting step for us and one that has generated a lot of activity and interest from our clients, but we're not restricted to anthropic. We have had conversations with other frontier model companies. Some of our clients are using other models. And we believe that our data sets will be relevant to the models. What's differentiating is the quality and the breadth of the data. And we certainly think that models will be differentiated, but all of them are going to require the data we provide. It's not just generative AI. We also have had a client for a very respected global insurance company that has asked us to build with their partnership and agentic AI underwriting platform. So this goes beyond generative AI to building agents that replicate the function in gathering data, assessing data and rating policies on a purely agentic basis. And we're excited to develop the intellectual capital and the intellectual property associated with that. But we've also talked about a client that has decided to pursue neuro-symbolic AI, now this is an important dimension of AI. It's actually existed for a while. The key differentiation here is that neurosymbolic, unlike generative AI, which is predicting the next token in response to a query. And so it was a predictive path, but in many ways, a random walk just as a stock price in a stock price random walk context, it's -- generative AI is not replicable and it's not auditable. Neurosymbolic was developed to apply neural network and machine learning approaches constrained by domain knowledge so that you could have a clearly determined reasoned outcome that was auditable and repeatable. And so we are working with a client to integrate our data sets with a neuro-symbolic model that we believe could certainly represent a very relevant AI technology for the industry and one which interestingly insurance regulators have already begun thinking through. And so finally, beyond that engagement, the other factor is the economic factor. We represent slightly over 30 basis points of industry U.S. net written premium and about 37 basis points of industry expenditures. Now the good news is that while it's small, it also has been growing, reflecting that increasing adoption of data and technology by the industry but there is clearly is substantial room to grow, and it speaks to the fact that for a relatively small component of their combined ratio or net written premium, we can deliver a lot of value if we can continue to bring that to bear as we have in the past on the functional efficiency with which they are handling underwriting, risk management and claims functions. So we feel we go into this with very distinct advantages, a historical level of trust and established centrality and very high-quality data sets. And all of that translates into a financial model that has some very compelling elements. As we talked about, it has very consistent and strong mid- to high single-digit organic constant currency revenue growth. We have strong operating leverage that drives improved EBITDA growth off of that 6% to 8% revenue growth. And from that, the ability through capital management to drive even higher EPS growth. We have a very strong balance sheet, and we have a very well-established disciplined capital returns methodology. The growth algorithm, as we described at our Investor Day that we anticipate over the next 3 years to continue to deliver 6% to 8% organic growth, of which pricing where we capture the value, both the value that we're delivering to the industry as well as underlying industry growth will deliver, we expect 350 to 450 basis points. That will be complemented by additional upsell and cross-sell where we have demonstrated an ability to deliver increased product sales to our growing clients. 100 to 150 basis points of new initiatives where we have new emerging technologies that we have an opportunity to penetrate the industry effectively, growth in new clients in adjacent areas such as intermediaries, and on the regulatory front and then offset by a historical level of attrition that comes from consolidation within the industry. So you can see that our growth model is one that's founded on value creation centered in pricing and expansion through both upsell and cross-sell of new initiatives and new clients. And then from a margin perspective standpoint, naturally within the system, our ability to make that investment at a relatively low OpEx and CapEx costs deliver strong operating leverage. We've enhanced that through efficiency gains by applying new technology like cloud computing as well as AI that we believe will improve our efficiency in data ingestion and analysis and then offset to some extent by portfolio impacts from some of our faster-growing businesses that are operating at lower margin, but we believe that, that's a good trade-off as those margins are actually improving as they scale as well. And we expect from that to continue to deliver between 25 to 75 basis points of margin expansion annually. And all of this is guided by a very clear and consistent capital allocation priorities. Our priority is organic investment, particularly in this environment where we have demonstrated an ability to invest in our data sets both their quality, the currency, the expansion of those data sets that our clients have been willing to pay us for because they see increased value in their utilization and the broadening of the applicability of those data sets. We look to invest in M&A where we see an established product, we can ramp much more effectively with our centrality and our scale and our position in the business. And where we don't see adequate returns from an M&A standpoint, capital is returned to our shareholders and we'll utilize to maintain our strong financial position from a levered standpoint. We generate a lot of strong core free cash flow, and we'll work to improve that through working capital efficiency. We have an investment-grade rating, and we have easily maintained a 2 to 3x to EBITDA target. So we're working from a very strong financial perspective and all of that supports our long-term financial targets of organic constant currency revenue growth of the 6% to 8% magnified through operating leverage and margin improvement for strong EBITDA and EPS growth and an attractive capital return with over 75% of free cash flow returned to shareholders. So hopefully, that gives you a sense of the very strong durable base that we have. The ongoing growth opportunity to deliver on the mission that we've had for decades, which is being a true utility to the global insurance industry where we can leverage those data sets and our scale to deliver strong incremental value to the industry. And all of that reflected in a financial model that is very durable, sustainable margins, strong organic growth and tremendous opportunities ahead of us. So I think that pretty much covers everything that I had, Andrew.

Great. Lee, if you don't mind, maybe we have a few more minutes. Just one kind of high-level question on the state of the industry as a whole. Can you talk about where kind of net industry premium growth looks, loss ratios? And maybe more holistically, your clients' willingness to spend on data and kind of the secular trend there?

Sure. Thanks, Andrew. So as I talked about earlier, the net written premium as a whole, obviously, a diverse set of risks. Across those businesses, we have hard markets, we have soft markets. And consequently, it tends to vary from year-to-year. We've done an analysis that we've shared previously that when we look at our organic growth in hard markets versus soft markets, our soft market performance, the growth rate has generally been 6.8%. And in hard markets, it's been 7.3%, so it's about a 0.5 percentage point of difference, both of them around to 7%. So it's an influence but it is not the primary driver of growth. It's the value creation that I've spoken about on where we can apply a technology to our data sets and deliver value for the industry and participate in that. Now as we are in certain soft markets, in hard markets, when everything is going well, there's a natural expansive element, which is great to have. But in soft markets where pricing risk becomes more importantly, our view is -- our experience has been that the data sets become more valuable and our clients tend to lean into that dynamic. And I think the second part of the question is how are our clients communicating to us about how they are thinking about utilizing our data. And I'll start with, I think, the strongest experience that we've had, which is we continue with our largest, our most sophisticated clients, the ones with the largest scale businesses, to see a consistent renewal trend of renewal of our multiyear agreements at or longer than the last renewal. This has been true for the last 12 months on all of our major renewals. And with price increases that reflect the increased value that our clients are experiencing from our most recent investments in our core lines reimagine effort. But we also have other data sets that we've been expanding that are new sources of growth. It isn't just the pricing element, but uptake on our excess and surplus capabilities, uptake on our catastrophic risk modeling new platform, Synergy Studio is something that our clients see incremental value in managing their risks more actively. And then finally, I think the thing that's been most promising so far this year, Andrew, is after a period of experimentation with AI the recognition from our clients that finding ways to tie this to our data sets and to their existing workflows is critical to them generating a return on investment. Now there are efficiencies that can be accomplished, but the ability to enhance that with broader operational and industry loss data, our experience at a micro level for those that have been experimenting with it are substantial, and I think that, that is going to propagate across the industry as we find ways to embed or as some of the frontier model companies as this technology diffuses into the insurance industry. And I think that will be a slow adoption process. The industry for both regulatory reasons and natural conservatism tends to be very cautious in how they adopt technology. But one thing I can guarantee our data will continue to be valuable in that process, and we'll continue to be a natural partner to the industry.

Great. Thank you very much, Lee, and thanks, everyone, for joining. We are moving to Adler for the breakout here in a few minutes. Thanks for joining us.