Podcast 57. Bob Reville, CEO of Praedicat: How to build a liability model

Twitter icon
Facebook icon
LinkedIn icon
Podcast

Bob Reville founded Praedicat in 2011 and the company has gone on to create one of the most credible models for understanding emerging risks.
The idea was born out of a research and development project, with Praedicat now using data science and AI to build products for some of the world’s largest insurers and global investors.
Praedicat recently completed its time in Lloyd's Lab cohort 3 and Matthew finds out how they are helping insurers and others manage large scale liability risks, the impact of regulation and the threat to insurers from a new wave of litigation issues.
The InsTech London podcast is supported by the Insurance Insider, who is offering our listeners a free copy. You can download the latest issue from insuranceinsider.com.

Listen here to podcast 57. It is also available on iTunes, Spotify and Podbean.

..................................................................................................

Transcript for this podcast

00:00 Bob Reville: And you're absolutely right. It's a much more compelling case to be able to go into an insurer and say, "We can help you double the size of your business, or more."

[music]

00:15 Matthew Grant: Welcome to the InsTech London podcast. Matthew Grant here and this week, we have my interview with Bob Reville, CEO and founder of Praedicat. Now, despite the explosion in insurance tech startups in the last few years, aside from cyber, there are still very few out there with credible offerings for tackling the intangible risks that today represent a far greater value of a company's assets, than physical assets.

00:40 MG: I've known Bob since Praedicat was founded in 2011, as a spin-out of the RAND organization. And since then, they've built out one of the most credible models for emerging risks, and are working with some of the largest insurers. They've got offices in California and in London. And at the time of this recording, Praedicat is in the third cohort of the Lloyd's Lab, and like everyone else that I've spoken to, it seems to be having a very successful time there. Bob's calling in by phone, so a bit of cracking in places, but some great insights into how to build a liability model.

[music]

01:17 MG: Bob, welcome to the InsTech London podcast. You are often over in London, but we're actually talking whilst you're in California, although I've got to say, it sounds like you're in the next door room. So it's the joys of the transatlantic communications.

01:31 BR: Well, it's a pleasure to be here. Thank you for arranging this.

01:33 MG: Great. Well, we've known each other, I think about eight years, probably even before you started Praedicat back in 2012. So, you're CEO. I want to hear a little bit about what you're doing with liability modelling, and you're also now on the cohort at Lloyd's, but could you just give a little bit of background, for people that don't know you and don't know Praedicat, about your own background, and then what you're doing at Praedicat.

01:56 BR: I am an economist and I used to work at the RAND Corporation, which is a non-profit think tank in California, and Praedicat emerged out of a R&D project that we did with RMS, which is how you and I met. And we were working on trying to figure out how to come up with ways to predict what might be the next asbestos, but to turn that into a system that would allow insurers to manage their aggregations around large scale liability risk. So sort of a cat modelling-like approach. And so, that's what Praedicat is.

02:37 BR: We've developed liability risk modelling. We've gone from, originally, doing cat modelling to doing somewhat more broad modelling around the science of liability risk. Meaning, what is... How is science informing future liability risk, and it's turned into products that we're selling not just to insurers, but also to global industrials.

03:03 MG: So specifically, what are you finding that your clients in the insurance industry have... What problems have they got that you're finding they're asking you to help them solve?

03:12 BR: Well initially, we thought that the problem was going to be only the issue of aggregation managements. So in the 2000s, again, when we were working with RMS, and we were working on terrorism insurance issues, and that's when I became aware of the whole approach of catastrophe modelling, and how it has helped insurers to avoid the insolvency that comes from huge correlation associated with natural catastrophes. And we were, at the same time, doing a lot of work on asbestos litigation in the United States, and started thinking about how all of the insolvency issues that had emerged out of asbestos might be able to be solved through a similar kind of modelling as existed on property cat.

04:01 BR: So our idea was that the problem that we were going to be solving was aggregation management, and all of the things that go along with that; maybe setting cat loads and things like that. But I think, over time, we've come to realize that the way in which the industry has responded to the potential threat of another asbestos, has caused them to, I'd say, be a bit risk-averse and to adopt an exclusionary mindset. So a tendency to, when new risks emerge, to try to avoid them rather than to try to cover them. And that's led to a large coverage gap so that today compared to, let's say, the 1970s, in 1973, 94% of US liability risk was covered by insurance, and by today it's fallen to about 55%, and all that just results from higher retentions, increased exclusions, lower limits relative to the growth in revenue, just in general, a contraction of the amount of coverage offered. And so, increasingly, we're focused on that problem rather than just the aggregation management problem.

05:21 MG: So it's really part of this whole theme around providing new information, and then giving underwriter confidence that that information is good enough to make decisions about. I guess what's particularly interesting about what you're doing, is it's actually opening up, or I guess reopening classes of business that they aren't writing or would have written before. But are you also finding, in terms of people's concerns about these areas, I mean it's a long time since people were impacted significantly by the asbestos losses, but is there a situation where unless there's been a really large recent loss, people tend to get a little bit either complacent, or they sort of forget these large losses happen, so you have to sort of re-educate them as well on the scale and types of losses that can be out there?

06:00 BR: Well, you know, that was definitely true in 2012, when we started. I'd say over the first few years after we started, we would talk about how we were trying to help the industry deal with a next asbestos-type litigation and some people said to us, "That's really not a problem anymore. We don't see that kind of litigation anymore in the United States. We're really worried about things like deep water horizon or a train wreck, or something like that." And so, others, of course, thought that it was just a matter of time.

06:00 BR: And as it's turned out, I think that the people who thought it was just a matter of time were right, 'cause today you're seeing emerging, a whole new wave of bodily injury mass litigation. You're seeing it with Roundup, and you're seeing it with Johnson & Johnson and talc, and you're seeing it with the opioids litigation which I think is potentially on a scale of an asbestos really, and you're seeing it with the PFAS litigation, repetitive head injuries. It's almost the '70s and '80s all over again.

07:10 MG: I think you said PFAS in there. Did I hear that correctly?

07:12 BR: Yeah, PFAS. That's perfluorinated and polyfluorinated chemicals. So this is things like Teflon and... So PFOA, PFOS and all the substitutes that were developed for it over the years, and they're now calling them the Forever Chemicals, which tells you all you need to know about the risk.

07:31 MG: Right.

07:32 BR: They don't go away.

07:33 MG: Yeah, got it. Okay, good. Another insurance acronym to add to the list, the PFOS. And so, I guess this is really interesting. So these are kind of emerging threats. How does that sort of work in terms of the point to which people shift from not taking it seriously, to get starting to taking it seriously but there still hasn't been a big loss yet? Do you see a kind of trigger point at which, suddenly, companies are worried, therefore there's insurance opportunities, and therefore insurers start to see opportunity in there, or is it more of a kind of slow change, an incremental change?

08:08 BR: Well, The nature of liability risk is that it tends to be slow and incremental. When you cover something, the losses actually probably won't be paid for five years or so, and if you cover something that... The way that it works in liability insurance where you are covering it on an occurrence basis, meaning when the loss occurs, rather than when the claim is made, it actually can be 20, 30 years before you pay on a policy year that you're writing right now. And so yes, everything tends to move relatively slowly.

08:40 BR: But that said, when you're seeing a large amount of losses, I mean the Monsanto-Bayer glyphosate loss is probably going to be pretty significant, and there are a number of other companies that are also selling that product. And talc is the same way. You've got the Johnson & Johnson loss but talc is a cosmetic ingredient that's used by a lot of other companies as well, so there's a lot of concern there. And then opioids. That one is definitely leading to a lot of concern for, not just the pharmaceutical companies, which tend not to be covered, but for the distributors and retailers. McKesson, AmerisourceBergen and the Rite Aid, and CVS, and the like. All of those are being sued right now, and are accruing defence costs, and where there's concern that there could be larger losses all around. So large-scale clause, I'm sorry, large-scale clash is definitely emerging as a risk in liability, for bodily injury-type or public health-type litigation.

09:56 MG: Are you focusing mostly on the sort of known risks, so all the things you categorized as emerging risks, seeing some trends, mitigation or are you also able to model, in some way, the unknowns that are kinda lurking in the background and not yet kinda made the headlines, but you're just starting to see trends, or they've got characteristics that could lend themself to become larger losses?

10:22 BR: Well, definitely not the known knowns, because that is not terribly useful to our clients who have already written the known knowns. What you want to try and do is to get the... Well, we actually... Because you're making a reference back to the old Rumsfeld, known knowns, known unknowns, unknown unknowns. And so, I think that people have always thought of liability as full of unknown unknowns, and our whole approach is to try to turn the unknown unknowns into known unknowns, using what the scientific literature is saying about what businesses are doing, and what products they're selling, and try and catch the risk at the earliest possible stage, when the scientific literature first starts to emerge, and then to, as those literatures evolve into what will eventually be a scientific evidence base, if it ends up in litigation, we are mining the literature progressively to try and understand all the exposure settings, and to understand the populations that are exposed, and turn that into, essentially, a simulated mass litigation. And so, for Roundup, for instance, and for talc, and all the PFAS chemicals, we were tracking them in 2013, when we created our first release of the software, and none of them were seeing real litigation at that time. And the whole point is to try to get at it early.

12:00 MG: Interesting. And I read in the website, you're scanning 30 million scientific journals and profiling 35,000 companies, so you must have got some pretty powerful analytics, to be able to extract that data and, I guess, also ways of actually extracting unstructured data from the journals as well.

12:17 BR: Yeah. So we are searching through all the published peer-reviewed scientific literature, looking for where there might be a chemical, or a product, or a business practice that scientists think might result in bodily injury, property damage, or environmental damage. So we start with that. We try and catch the journal articles at the earliest possible stage, and then we connect the articles into literatures, and then track those literatures as they become, essentially, scientific evidence.

12:49 BR: And then at the same time, as we start to mine those literatures to describe the industrial footprint that is exposed to that risk, we turn that into company profiles, for the companies that are in that exposure footprint. And then both of those modelling efforts, the exposure modelling effort and the... Well, so I should say it's the scientific evidence modelling effort, and then the company exposure modelling effort. We start by doing, essentially, by hand with experts, curating data and then over time we turn all of that into training data, which feeds machine-learning algorithms, which allows us to get up to the millions of journal articles, and the tens of thousands of companies.

13:42 MG: So a lot of companies claim to have AI. It sounds like, you [chuckle] actually... You do have AI, and you need it, and you're proving that you can use it. I just... I'm kinda interested, just thinking through practically how an underwriter would use your tools. So are they starting off with a company they're looking to insure, and they enter that company into the model, and that then comes out with a, sort of, risk scenarios and pricing? I'm trying to realize, is it analogous to how catastrophe models work? Or is it just something different you have to do to get to the result?

14:15 BR: No, it's exactly that. So you would have an underwriter, receives a submission or a renewal. And you would log in and find that company, and then you would see, first of all, out of the 250 agents that have literatures that are large enough, that we're tracking them as potential large-scale tail risks for insurers. You'd find out, Is this company exposed to none of those? To 50 of those? And then, all of that is turned into information that we quantify in terms of exceedance probability curves. And so you could get PMLs and TVaR. And we also turned it into time dimension risks. So it's not just... So the whole thing about liability is unlike in the property environment, where you write something and then if there's going to be a loss that year. For us, you find a company, they're exposed to, let's say, a round-up and we'll tell you not just that there's some risk that there would be litigation for this policy year, but that there could be litigation in future policy years.

15:32 BR: Or if you write this year, there could be litigation that would be covered this year that might not emerge for a few years, or for that matter, if you have an insurer that has... You've been writing in the same company for the last 20 years, that's one of the... That's the real hard problem in liability insurance. The stacking problem, this is where you actually can have 20 policy years, that would all be activated, because you had a company that was exposing the public to some chemical for 20 years, and every policy year could potentially end up paying out. So there's all this is... So, to get to your point though, you'd go in and you'd be able to see, not just the losses that could happen this year, but the full-time dimension of the loss that could emerge.

16:22 MG: Yeah, and it's fascinating. So yeah, it's on another dimension versus, I guess for want of a better word, the kind of cleaner cat loss it might be, I'm not quite sure what the reverse is, of the gift that keeps on giving. But it certainly sounds like the sort of the pain that keeps on... You'd think if you were a liability underwriter. What about regulation? Because that's certainly another area, as you know, regulation often drives adoption. Are there other similar regulations in the liability space, or as sort of rating requirements for the companies.

16:50 BR: There is a growing amount of interest in the casualty aggregation risk. So, for instance right now, the PRA has an initiative around this, where they're trying to understand how much exposure there is to cast the aggregation, and then to think about how to encourage companies to manage it better. Also, the rating agencies are getting pretty focused on this. AM Best is beginning to ask for scenarios. The Bermuda Monetary... Oh, so down the regulatory side, the Bermuda Monetary Authority has started asking for liability scenarios, and SMPs are getting very interested in this. I think it's early days still, about what concretely will emerge as the regulatory framework, but there is a lot of interest in it. I think that after you had solvency too, the first wave of regulation that came out of that was in property, and they... And historically, in fact, there probably has been more insolvencies associated with liability and reserve deficiency and liability, than there has been associated with NatCat. And a lot of that has to do with aggregation risk, and so there is definitely going to be some sort of a regulatory framework that comes out of it. But it's a harder problem in some ways, because of this time dimension issue, and so exactly how it's going to work, isn't clear yet. But it's these competing approaches that are emerging.

18:26 MG: Yeah, it's certainly a big big-business driver in terms of adoption, what I expect is that it becomes clearer, and more of a requirement. And just talking about business adoption. I mean you've got... You're so very well-known, or you're certainly well-known in this particular space, and there don't seem to be many other companies out there doing something similar. Is there a big barrier to adoption? We talked about the 30 million journals you go through, and the 35,000 companies, so clearly there's something around the, just the rigour, and depth, and processing power of the analytics. But you... Why is it so difficult to get these things to a credible point as a model?

19:00 BR: I think that the idea of doing something that is exposure-based and forward-looking which is what we try to do. Forward-looking meaning, you can't just model previous mass litigation events, because that is not interesting. Because those products have been removed from the market, and those companies are out of business. And instead, you've got to try and figure out some approach to do something forward-looking. And then also, turning that into company data. There's been multiple years of investment by Praedicat, it's... And technological innovation. We've had to, as you noted, develop an AI to get this to scale. So I think that is a bit of a barrier to competition. But that said, we do have competitors. They take different approaches, and the brokers also are involved in trying to model and look at the exposure in this space. So no one's doing it quite the way we're doing it, but there are others.

20:07 MG: Yeah, well that’s healthy. It's always good to know there's other competitors out there, otherwise you sometimes wonder why you do it.

20:13 BR: Yeah, exactly.

20:14 MG: And can you talk about some of the clients you're working with?

20:16 BR: Yeah. Well, we're working with 7 of the 10 largest liability insurers, and we work with about 18 insurers and reinsurers, and we are now also working with some global industrials. So by the end of the year, actually, we'll be at about 10 global industrials. So these are large chemical companies that we're also working with, and that area is pretty fast-growing, so I actually expect next year we might end up, maybe about mid-year, having more global industrial clients than insurance and reinsurance clients, which will be interesting.

20:56 MG: Are they coming to this because they're looking to change or reduce their insurance purchase, or is there something different that's driving their interest?

21:05 BR: Well, this AI as you call it, that we have developed, this ability to read through scientific journals. If you're a chemical company and you've got, let's say, dozens or even hundreds of products, and you need to track the science that is looking at whether or not your products might be causing bodily injury or environmental damage, and it turns out to be a large scale issue that also has regulatory concern completely outside of buying insurance. Now, we do take an insurance-focused approach to it, which makes us different than a typical product stewardship. That's what they call it, product stewardship. On a typical product stewardship vendor, we're much more quantitative in the approach that we take, and more focused on science, less focused on regulation. And we do expect that that, eventually, will be something that will help facilitate better, more efficient insurance buying, but that's really not the pitch that we are using at the moment. It's more just about, How do you stay on top of the risk? And how you make sure that you are doing the right thing by your customers, and the general public, and your workers.

22:21 MG: Yeah, it's a pretty powerful business model. There's very few companies that I can think of, and if done it's always been a challenge in the catastrophe modelling world to basically service the end client, who so often wants to, if it's buying insurance, they kind of want to buy it through conventional means. So to be able to help them understand that risk, I guess like in your case, understand and address the regulation issues, but also be working with the insurance. Ultimately, you get to a very strong virtuous cycle there, where the risk is assessed the same way, therefore the information is consistent and ultimately everybody wins, 'cause they're understanding the risk, and reducing it, and pricing it more effectively.

23:00 BR: Yeah, totally. I think part of the thing with liability is that, liability is the losses are not just what you end up paying in defence costs and indemnity. If there is a mass litigation against the company, then there's reputation risk, there is products that get removed from the market, so there's loss of revenue. And no one would say, right now, that the only cost that they're is bearing for the round-up litigation is the cost that they're paying in defence costs. It's much larger than that. It's really existential. So the fact that so much of the loss is not just what's covered by liability insurance, makes it something that these companies have entire functions built around trying to understand the risk, and then they are also buying insurance.

23:52 BR: I think that that insurance buying is... The liability insurance is a pretty blunt instrument at this point, and really is not capturing the texture of the risk for particularly large companies, and there's a lot of scope for product innovation to be able to better help the large corporations to manage this product stewardship risk. A lot of costs don't have insurance associated with right now that are, associated with trying to do the right thing in ways that will, if not done, later result in liability. So I'd say there's a huge scope for expanding the coverage available to deal with product stewardship. And that's really where I see Praedicat making a big difference in the long run, is helping to facilitate better coverage for the set of risks that are associated with liability, but for which liability is only one small piece of it.

25:00 MG: Yeah. It definitely seems to be a theme now, I've been hearing it in lots of different places, where companies are looking at their total balance-sheet risk and combining what was conventionally thought of as being insurance, but recognizing a lot of losses turn out that they're actually outside of insurance or are uninsurable, but ultimately they need to figure out where are the really big catastrophic events that could happen, and then have different types of protection in place, that may be different than the previous type of indemnity protection that has been bought so far. So it's very interesting to hear that's coming through on your side as well. Maybe can you just talk a little bit about how you, sort of, access your clients? Being based in Southern California, it takes you quite a long way from where the world of insurance is, and even further of reinsurance. And you just recently joined the latest cohort in the Lloyd's Lab so, How do you sort of find that balance, and what motivated you to join the Lab?

25:55 BR: It's the weather, really, that keeps me in Los Angeles. It's not the proximity to clients. We're very excited about being part of cohort three because it allows us, definitely, to integrate more with the Lloyd's market, with London insurers in general. It's very much a face-to-face culture there still, and so building that office out and then getting involved in the Lloyd's Lab, I think is going to be the key to help having people get comfortable with our modelling.

26:28 MG: Yeah, face-to-face but also, yeah, people take a while to understand new risks, and yeah, I think there's that attention in Lloyd's. It's very good at being a market for the risks that are really difficult to place elsewhere. But people... I can think of one writer who said, "Rightly are a bit like this, they're a little bit sceptical about new ways of analyzing risk." And so convincing people that the new technology is good enough to go and put some pretty significant lines down, is always going to be the challenge. One of the interesting things, and I don't know if you've specifically come across this, but with the new product innovation facility they've launched in Lloyd's, which is intended to provide access to capital for, initially anyway, some sort of test cases.

27:10 MG: And it isn't necessarily intended to kind of bet the business on it, but to experiment with some of these, and that seems to be pretty well fitted into what you're doing in some of the areas you're looking at, where you can start to write some lines, use some creative ways, such as indices and other sort of non-standard ways of pricing the risk, or triggering a payment. But are you doing much in that kind of area, or anything else you're seeing at, at Lloyd's that is sort of opening now... Opening it up a bit more now than the challenges you've had trying to engage before?

27:39 BR: Yeah. One of the projects that we're working on in the Lloyd's Lab, is the development of a new-named peril liability insurance product. So the project is that we are working with our mentors, and also working with others in the market, to try and scope what would be a new product that could be offered out of the Lloyd's market that covers tail risk in liability for a large-scale mass litigation on a named peril basis. And that would involve getting people comfortable with the model, and also understanding the size of the market, the companies that are exposed, why they're not adequately insured today, putting all of that into a business plan, designing the actual product, instrument, all of this. And then helping to have that also feed into a workflow in our software. All of that together is this business plan that we're developing, and that's one of our key initiatives. And I think it's going to be... I think it's going to be great. This idea... There's a lot of interest in the idea that liability, which is today almost entirely, I don't know on an all-perils basis.

29:00 BR: But what that all-perils basis does is it ends up leading to a lack of coverage for things that are excluded from the... It's all perils with exclusions. So there's lack of coverage for the excluded risks, and then there also is unwillingness to offer as much limit, because they don't understand exactly everything they're covering. The constant refrain you get from underwriters is, "I wrote this company thinking this was the risk, and then lo and behold, out of nowhere, this is the loss that emerges." So the more we can anticipate where there might be future mass litigation using science, and then cover the largest potential losses on a named-peril basis. The feedback we're getting from the market is that's well suited for Lloyd's, and it's about getting everybody comfortable with the concept, and then we're hoping that that leads to the launch of an entire new approach.

30:01 MG: Yeah, no, that makes total sense. So a named peril might, for example, the opioids as one of the named perils you might... Someone might buy cover for.

30:09 BR: Right. Or ideally, if you had seen opioids before the fire started, before the house started burning, that would have been... That and a set of other named perils, as an entire portfolio. Our idea is, you don't want to just be writing named perils for round-up or for opioids. But you want to be able to think where might there be 25 other public health issues that look like opioids, out of which, maybe three of them will result in litigation in the future. Or where might there be 50 other pesticides, and herbicides, or other agricultural chemicals, and any one of them could be the next round-up, but most likely, most of them won't. But they all have science associated with them, and they're all quantifiable.

31:01 BR: And they all can result in losses for certain companies that are above the level of their insurance today, or are directly excluded. And so being able to understand all that, and then cover it under a named-peril basis, will get people comfortable with offering more limit above what's currently on an all-perils basis. Or offering coverage for things like electromagnetic field today, which science isn't really all that supportive of the idea that there's going to be a lot of litigation in the future over cellphones causing brain tumours. But it's, largely it is covered today, despite the fact that science has moved away from it. So being able to encourage that kind of innovation, is what we're spending most of our time doing these days.

31:46 MG: As you're talking, I'm sort of thinking you must be a bit like a doctor in A&E, you're seeing all these horrible things that can happen, but that's kind of part of the day job, and you just have to keep going and hope that you can play a role, or Praedicat can play a role in highlighting the risks, and therefore encourage risk mitigation for those risks, yeah, despite the fact that, ultimately, they can become fairly unpleasant outcomes and lots [chuckle] of loss of costs associated with them.

32:11 BR: Well, a lot of stuff we do though is debunking, at the same time. So things like Diet Coke. A lot of people think, "Oh, Diet Coke is a... Going to kill me." And at the same time, when we look at the science around the artificial sweeteners in Diet Coke, it looks pretty benign to us, and so we don't worry about it, and over time... While initially you... When you're first exposed to the stuff that we do, everybody here has this initial reaction of, "Oh my God, there's so much horrible things in food, and the environment, and... " Particularly if you have young kids, a lot of us here have young kids, and so you... A lot of science these days is around how environmental chemicals and other types of products are causing developmental issues. So you get really nervous, but then over time, you kind of realize that, first of all, that you really just gotta think about the world like an underwriter should think about the world. That you can take on almost all these risks. You just have to not have too much of it, so you avoid your aggregations. And then also, you get to think... To find that the science around things like coffee, which is what fuels all of us here, is increasingly positive. It actually has health benefits, and when they used to think, it caused cancer, all the science no longer supports that.

33:39 MG: You've been around for awhile now. What's your sort of perception of what's happening in this whole insurtech area?

33:45 BR: In the insurtech world, there is just as much of a pressure almost, for us as an insurtech company to be an insurer, as there is for the insurers to become tech companies. It's this convergence between tech and insurance that pushes us in both directions, and it's pretty interesting. I don't know where it's all going to shake out for us, but it's definitely interesting.

34:08 MG: It's really interesting, which companies are doing things like you, where you're actually offering new solutions to new risks or, to pull your earlier point about how there's been a shift in what people are covering for liability. Yeah, all the risks that people are no longer comfortable covering at. You've got a much more compelling story if you can go in to an insurer and say, "I've got some tools that can help you underwrite more business and grow your revenue." Rather than, "I want to tell you all the risks that are associated with what you're already writing, and you could have spent more money to understand what you are doing better, and all the costs, and difficulties related to that." So yeah, I think there's definitely the companies that are going to be amongst the more successful going forward, if they can collaborate with the market, and particularly, everyone that is talking about the intangible risks, and you're a big part of being able to sort of figure out how to analyze these intangible risks.

34:55 BR: Just to follow up on that, I mean when I described the fact that we've become more focused on addressing this coverage gap, rather than on just helping manage aggregation risk. That coverage gap is about 75 billion dollars annually, that's uncovered by in-liability insurance in the United States. And that's a... Just try to get back to where the insurance industry was in the 1970s, but to do it right, to do it in a way where aggregations are managed, a 75 billion dollar growth opportunity is pretty huge. And that's to say nothing about the other risks that are not even covered by liability at all, the other sort of product stewardship risks I was describing earlier. And you're absolutely right, it's a much more compelling case to be able to go in to an insurer and say, "We can help you double the size of your business or more, and at the same time, help you to do it in a way that's sustainably profitable, unlike the way liability insurance operates today." That's why we've become more focused on the coverage gap than just on the aggregation management, because it's definitely a much more compelling pitch.

36:17 MG: Well Bob, it's been great to carve out some time. And I hope we will see each other face-to-face when you're over in the UK, at some point. And it'd be good to check-in at some point in the future. See how things are going once you come out to the Lab and yeah, let's see how that's worked for you.

36:29 BR: Sounds great. Definitely, I'll let you know next time I'm going to be there.

[music]

36:36 MG: You can find more episodes, featuring others in the Lloyd's Labs in earlier recordings, and more to come. Now the InsTech London podcast continue to be supported by Insurance Insider, and we're delighted to offer you a free issue. The details are in the episode notes. If you want to find out more about everything we're up to at InsTech London, then the best thing is take a look at us on www.instech.london and feel free to contact me via LinkedIn or on matthew@instech.london to tell us what's on your mind.

Listen here to podcast 57. It is also available on iTunes, Spotify and Podbean.