Verisk Analytics, Inc. (VRSK) Earnings Call Transcript & Summary

Hi. It's Andrew Steinerman at JPMorgan Equity Research Department, and I'm a business and information services analyst, and this is the Ultimate Services Investor Conference, our annual forum where we bring together the best business and information services companies and investors. This is the Verisk fireside chat. It's going to be lively. We're interviewing Scott Stephenson. He's been CEO since 2013. He's with Verisk for over 20 years. And also, I'm just going to mention on the website is our information services data book. So on the same digital book that you with watch this fireside chat, you could actually pull off our just published quarterly info services [indiscernible]. Scott, welcome. Thank you for joining us.

Good to be here.

Could you just talk a little bit about capital allocation and ROIC? It seems like this is a particularly timely juncture for Verisk to be reflective about that.

Yes. Well, you said it. We always have tremendous focus upon trying to run our business with -- in a balanced way. And where we always start actually is what do our customers need, what can we do productively on their behalf to bring new innovation to bear into their worlds. It's rewarding for them. It's rewarding for us. But right alongside that, we're always asking questions of capital efficiency, what does it cost to generate that next solution on behalf of a customer? How much of a difference does it make for them? And so we've always got all of those considerations at work. And it's -- and that applies not only to the next dollar of capital we invest but also the last dollar of capital we invest in. And so we are at a moment where we are taking stock of the fact that we have attempted to take the same playbook, which has been at work inside of our insurance vertical for a very long period of time. And we have, over the years, attempted to sort of [indiscernible] that playbook into other vertical markets. And we're asking the question, how productive has that been? And how productive is it likely to be. And -- but it's -- with respect to it being about capital discipline, it's no different than it's ever been. And we're always asking questions about the efficiency of capital in the 75% of what we are, which is insurance as well as the 25% of us, which is not insurance.

Yes. I think it was around 2017 when ROIC became a much sharper discipline. It definitely became part of a company framework and not just for investors, but throughout the organization, ROIC, the way I calculate it in our information services data book, Verisk has been around 13%, 14%, really kind of steady over the last couple of years. So my question...

It's actually gone up materially over the last 3 years.

Okay. We might calculate it differently in my data book, as a footnote on my calculation. So just state what ROIC is under the Verisk calculation and what should we hold you accountable for in terms of goals and ROIC.

You should hold us accountable for our enterprise level ROIC continuing to rise as it has over the last several years. Tens of basis points -- dozens of basis points of gain in any given year is pretty good. And if we can compound that steadily over long periods of time, I think that our shareholders will end up feeling very rewarded. And I'll just note here also that one of the things that our Board has been about recently has been looking at ways that we can reflect capital discipline into our program of compensation even more directly. We have [indiscernible]. So it's an abiding part of [indiscernible].

Right. So just a question on ROIC. We'll surely figure out the difference. So on my Page 12 of the information services data book, I include intangibles. Do you include intangibles when thinking about ROIC?

I actually want to make sure I'm technically correct in the answer. So I'm going to hold off on that, Andrew.

Okay. Sorry, sorry. So just let's pivot that same question. Do you want us to hold you accountable for the ROIC on acquisitions collectively?

Definitely.

Okay. Okay. Perfect. One of the things that I think you've just talked more about recently is kind of the importance of hiring the right people. You're surely -- obviously, I heard Jamie Diamond talk about the same subject. But for you personally, I just have heard it more from you higher than retaining the right data scientists, right people? Is there something like right now that's different, or hey, it's a really competitive labor market, and we have a great proposition to offer?

Yes. There are 2 things, which, to me, are different today than they were 18, 24 months ago. One is the competition for, I'll call it, technical talent is just more intense than it ever was. And I will say that we have felt rewarded for our own efforts to build our own pipelines kind of with early people early in their careers, bringing them in and training them to be cloud architecture, data analysts in the way that we want data analysts rather than going and trying to hire them away from other companies but really going early and deep into the pipelines. We've also diversified the places where we have our technical talent. We actually have a large footprint in India. We also have a large and growing footprint in Eastern Europe where there's also a good deal of technical talent available. So that's one thing. But to your question, the change is the intensity of the competition has gone up. The other thing, I think, it's actually probably even more profound with respect to talent is, I'll use Verisk as an example. We've got an order of magnitude 2,000 people [indiscernible] supervise the work of other people about order of magnitude 2,000. We have a critical dependence on what they're doing. The way that they organize their teams in order to create not only productivity, but retention, employees that like being at Verisk -- our employees do like being at Verisk. But a lot of what a person experiences at a company is their direct supervision and the team that is right around them. So we have a big dependence on those 2,000-ish people to create great team environments, et cetera. And the thing which has changed is the balance of on-premise and remote work is different, has been and is going to remain in that way. And almost nobody that is supervisory in the world today grew up in a world that was this hybrid as the world that we're going to be in. So we need those people to be very talented, to stay on the word talent. As it relates to how do you run teams in the new normal. That is super important.

Got it. Okay. Let's talk about 7. As you know, 7-plus organic revenue growth as part of your multiyear framework. The last time we sort from the company formally with the 2018 Analyst Day. My question is, does 7 -- and again, I don't know if you emphasized the plus, but I've always thought of it as 7-plus, still maintain its relevance as we think about the next few years ahead. And why is 7 the right number? Or should I be thinking plus?

Absolutely. We are sticking by our long-term model, and we're ambitious. So I don't resist your plus. You can just look at the 75-ish percent of us, which is insurance. And you could see that over a very long period of time, that's been a really good goal. And so we remain on that, basically. And so many good things happen. Why 7? Because so many good things happen in terms of your incremental margins when you're there or above.

Totally. And in the past, you've said, hey, we believe that all 3 of our verticals could grow that 7. I think it's fair to say that Energy and Financial Services hasn't, let's say, in recent years, grown back 7. My question is, firstly, on Insurance, do you feel like Insurance is more likely to be above that 7, emphasizing the plus a little bit more? And then secondly, let's jump into just Energy. Let's not talk about Argus at the moment. Do you still see a trajectory where WoodMac and the general data analytics of energy and research that you provide could be a 7 also.

Yes. So starting with Insurance, I mean, even if you look at the most recent quarter, you see the plus. And it is the right way to be -- it's the right level of ambition for that business. And just an observation about the macro environment would be that relative to, say, 15 years ago, the cycle has been completely squeezed out of the P&C, the insurance world.

You mean good, right? You mean it's a more stable industry.

Well, yes, but you don't have the actual trend line from [indiscernible], but there's no volatility. The reason that I cite that is our growth rates are being recorded against an industry background, which is basically a flat line. So we've been turning in the performance we've been turning in without this -- a rising tide in terms of the underlying marketplace. So we can -- we have and can grow that business even if the underlying market -- it's stable, but it doesn't grow very much.

But wait. Just [ let us pause ]. Is that better? Like to me, I always sort of dig like attractively the net rating premiums and seeing the volatility because volatility is ups and downs. And so like is it easier to sell to the end market that is just kind of, let's say, over the long term more stable?

It's constructive kind of more versus not. I mean I'd have to visualize the alternative that we're talking about. But yes, I'll tell you the 2 underlying conditions in Insurance. One is very stable, as we were just saying. And the other is I spent time with dozens of CEOs of insurance companies on an annual -- I go camp out in their offices and talk to CEOs. Every one of them thinks that their business is a growth business. Every one of them. And so that's also constructive in terms of what we do because they're ambitious data analytically because they want and expect to grow. Energy is a good business. The mix of what we serve has changed. We used to have a lot of exposure to upstream hydrocarbons. We have much less. But if you look at all points in the Energy ecosystem, that world is very dynamic. It wants to become more data analytic. We have very good assets, and that business is going to do well going forward.

That makes total sense to me. Something that really caught my ear the last conference call was you said we're soon going to turn off the last mainframe, like I think this is the first time I ever heard you talk and say soon. Like we've been talking about that for a long time. Could you just help me -- and maybe broaden that out, talk about both your mainframes and your data centers together. When is soon? And what do you mean by a turn off?

Yes. So pick those apart. So we've already taken a step, which is essentially variabilize the cost of that, which we do in the mainframe configuration, because we've essentially outsourced what used to be our premise-based [ new train ] configuration. So there are companies in the world that will provide you back your mainframe computing as a service, and they variabilize it according to the compute load that you put on it. So we've already done that. And relative to the number of cycles that we used to have, they're down materially. Think of it as sort of order of magnitude, 75%. So we're going to -- in this variabilized way, we're now going to -- we're going to squeeze out the last 25%. So that's what we're doing with the mainframe. With respect to data centers, we have 2 power data centers in the United States and one in Europe. And essentially, within kind of a year plus time frame, because we've migrated so many things to the cloud already, we'll be able to close those data centers.

Okay. And could you talk about what the benefits will be when you think about kind of the P&L when we do all that?

Yes. If you look at the amount of money that we will save when we have done that, you're talking about -- relative to the cost of our overall computing infrastructure, you're talking about material numbers of basis points in terms of the cost of computing [indiscernible] in our company.

Okay. Perfect. Something that also sort of curiously kind of stands out to me going back to the 2018 Analyst Day was the double-digit EPS growth. And I know this is a multiyear framework. I just want to make sure that, that's reported double-digit EPS growth. And over what period of time should we measure the company when thinking about double-digit reported EPS growth?

Yes. I mean that as the model is the model. And our business, it's -- I mean it's pretty easy to forget, but we have gone through a pandemic, but kind of a normal -- on a normalized environment, I would expect, by and large, that our long-term model applies in the short term.

Okay. And maybe this is just kind of a directional question. But to achieve that double-digit EPS growth and then we started with 7-plus organic revenue growth, that has to include kind of reported margin expansion over time, right?

Absolutely.

Okay. Perfectly. Okay. That's great. Now maybe talk a little bit about the Argus journey. When I wrote up the note, I think it was 2012 when you acquired Argus. Also a unique data asset, obviously, something essential to the credit card issuers, full coverage. I would have also thought that this was an asset with a lot of leverage to expand into adjacent markets. So could you just tell us about what worked, what didn't work, where are we on that journey?

Sure. So when we moved into it, our hope was that relative to what we were starting with, we could grow at 10x. And over the time period, we've grown at about 3x. What have been some of the puts and the takes in terms of the growth rate? One of them has been, in a Dodd-Frank world, the degree to which our banking customers are able to adopt and commercialize the models that we build out of this wonderful data set had sort of been -- the opportunity has been not eliminated, but constrained because it used to be that we would generate, and we can still generate a model, which is basically -- and all of -- you're working with data, which are anonymized to the background. But essentially, we can get to a view of any individual across all of their credit cards, essentially the share of wallet at each of those cards [indiscernible], right? It's a really neat analytic. It used to be that banks could use that and just use it. Now if they want to actually use that inside of the way that they're running their business, they have to make like a 100-page filing to justify the analytics. So we've had to sort of extract signals from what we were doing but not presented in the same way. And it kind of caused a reset in terms of what our customers were doing when the regulatory burden went up. So that's one thing.

And also give a time frame to that, when did that happen?

Well, the actual -- I'm going to say...

The impact.

Yes. I'm going to say that was like '18, '19, somewhere in there. Another thing is we've had very high expectations for taking the signals on spending and turning them into a view of what's the utility of the advertising and promotion that's being done, and you [ watch the ] spending results. And that's still a really good business, and that's still a part of the business plan. But in a world, which has gotten much more concerned about privacy, essentially, we've had to kind of replatform what we do before sort of going higher. And it's also been the case that some of the sort of name brands, digital media, social media kinds of companies that were coming into the space in a big way sort of kind of hit the pause button. They're not out of it, but they kind of hit the pause button. But now replatform for greater privacy that, that business is poised to take the next step up. So that's another thing that happened in that space. It's a great business, unique assets, loved by our customers, all of that. But my hope was 10x, and it's been more like 3x.

Yes. Got it. Let's get back to Insurance. One of the things that's very clear that's evolved in your thinking about Insurance, which is obviously your best and biggest data analytics business, is to develop software that enables the data. And obviously, LightSpeed has been an example -- a big, successful example within Insurance, particularly automotive Insurance. Could you give us other examples of kind of developed software in Insurance or middleware that you have on? Or are -- of course, I'd love to hear -- you're in the process of doing? Like why are you so excited about kind of marrying data with developed software in Insurance?

Yes. So some other examples in Insurance would be Touchstone, which relates to our [indiscernible]

It's inherently Insurance. Yes, yes.

Extreme event modeling. And that's a really -- excuse me, a very important case because if you look at the competition in extreme event modeling over the last 5, 10 years, we have always had, we feel, superior science, and we feel like we've always been credited with that and superior models. But the substantial gains that we have made in that space over the last 5 to 10 years, you can attribute a lot of it to the software. And just to make sure that all of our friends who are on this call get this clearly, when you increase the software intensity of what you're doing, starting from a fundamentally content-rich analytically deep place, which is where we started from, two things happened. One is the attachment point with the customer becomes much stickier, one, but the analytics software itself carries value. So you have a pricing opportunity by virtue of now marrying your content and wonderful signals that come out of your analysis with the software. So Touchstone, LightSpeed, ClaimSearch 2.0, exact analysis, some things that we are building with respect to properties, just a whole host of examples, but it's a very major movement inside of our company.

Right, right. But you -- to say a little bit more about properties. Obviously, for insurers, you already have the key information about buildings. How will software enable that content better?

Well, there's a couple of things. We have a lot of data about properties, but there's more data we could get from our customers. And so the software will enable that. That's one. And then what we turn around -- and then the solution that we provide them will be inherently more consumable for them going back the other way.

Right. That makes total sense. And how much lift do you feel like you can get when you already have a proprietary data set, again, like building information and then you add a software on top of it? Like when you put those 2 together, how much lift, how much value creation are you providing? And how much are you able to monetize?

I'd say the effect, all else equal, and now it's the same content, it's the same analyzed signals, but what you've done is you've surrounded it with software, 10% to 30%.

Right. And that's your monetization, right?

That's our monetization, right.

Right. And then the little -- other part of that question is if you're getting 10% to 30% more because you're putting the software on top of the data, how much lift in value creation do you think the customer is getting? If you think of 10% to 30%, I'm sure the customer must be getting more than 10% to 30%.

Much more. Yes, multiples of that.

Right. That's what I would think.

Yes, multiples of that. I mean, basically, we find that customer buying of something new that we bring to them ignites when they -- when the benefit they get relative to the cost is about 10:1.

10 to -- wow.

Yes. I mean one of the things about our business is we -- our customers are sophisticated, and they're [indiscernible] entities. So they are very -- they do a lot of diligence before they decide to use a new solution, ours or anybody else's. And so the value proposition has to be evident.

Right. I'm a little surprised by the 10:1. If you can ask me a guess, I was going to guess 5 to 7x. So my question is, is that for a brand-new product 10:1 as an aspiration? Or even when you're talking like a software enablement on top of data, is it still like that ambition of let's provide 10:1 to the client?

No. I was summarizing a lot of experience over a long period of time with the 10:1. But you're right, if they're already sort of in our world making use of our content and now we're amplifying it with software, yes, the cycle of adoption would be faster.

That makes total sense to me as well. Could you just tell us -- I know you're going to say we've said what we said, but just go over a little bit of the process ahead with the portfolio review. It definitely sounds to me more about potential divestiture or divestitures than any large acquisition. And how are you going to evaluate potential divestiture or divestitures?

Right. So I'll just comment, first of all, why are we even engaged in this? The reason we're engaged in this is for a very long time, having laddered up our Insurance franchise, if you go back 20 years, it has mid-teen margins and was growing very little. And so we -- but it's become what it's become. So we know what that journey looks like. And the dream has been, can we do that again in some of the [indiscernible]? So we picked several verticals where there is 2 conditions exist. Transaction volumes are [ great and ], the U.S. P&C industry. And secondly, proprietary data is in the hands of a third party like us. And we've said let's see if we can recreate it. And I'm summarizing a lot when I say we've made platforming acquisitions in the 4 verticals that we have found, and they're the only 4 verticals that have those characteristics. We made platforming acquisitions, and then we took account of what was the macro environment doing, what was the competitive environment doing and what was the actual performance of what were the returns on capital. And against those 4 verticals -- so we have moved into all 4 of them. 2 of them to date, we have moved out of. And we made good returns for our shareholders in doing it because these data analytics businesses, their franchises. They don't wither up and go away. These are good businesses. But we're always going to ask the question, what is the highest and best for a Verisk shareholder? And given the predominance of what we are in insurance, inevitably, anything which is an insurance is going to get compared to what is insurance. And I'll just observe that the portfolio of businesses we have today, last year generated about 40% TSR. This year, we haven't traded as well. Well, you have [indiscernible] listen to that, that's a signal, and that's a signal that we've heard and that we're listening to. But fundamentally, the analysis is not just the next dollar of capital, but the last dollar of capital. Where is it right now? What could these assets be worth to someone else maybe? And what do we think are the fundamental prospects? One of the reasons why we're trying to be very deliberate about it is, I'll just remind everybody, we've been going through a pandemic, highly disruptive. And so we've really been trying to read the signals of what our business is doing as we move into bottom of and now hopefully pulling out of the pandemic. We're reading all those signals in order to have the very best view of what's the likely course of these businesses going forward. That gets put up against what are the alternatives. But there isn't a thought here about, well, okay, if we liberated a lot of capital, for example, what's the next vertical? We identified 4 verticals that we thought were interesting and maybe we could recreate what we have been building and have built in the insurance industry. And the report is of the 4, 2 of them we exited because the environment has changed, and the other 2 we're examining closely.

That makes total sense. Why don't we just end on one question? So think about next Ultimate Services Investor Conference, November 22, of course, will be in person, my question is, what do you think you and I will be talking about Verisk that we're just not talking about today? Obviously, investors are very focused on this kind of portfolio of your question. What do you think you and I will be talking about a year from now that we're not talking about in the time that we just had together?

So I mean I'll be interested in your answer to that when we get there also. So I'm -- because you're always lively with what's the next turn of the wheel. But some of the things we might talk about would be Verisk playing in the global landscape, in Insurance, that could be one. Another one could be the relationship between all of the climate science that we've got, which we applied today inside of the Insurance vertical, what is the relatedness of that to topics of sustainability and resiliency, and whether and how that crosses over into -- beyond insurers, which is the majority of our book of business today, into the corporate world. Those might be a couple of things. But I really -- we'll see. We'll see.

Okay. That was perfect. Thank you, Scott. We always appreciate the dialogue.

Me, too. Thank you. Thank you, Andrew. [indiscernible]

All the best.

Bye now.