top of page
Search
Writer's pictureMark Geoghegan

Transcript: Ep145 Adrian Jones

Updated: Oct 29



Mark Geoghegan Adrian, welcome to the voice of insurance. Adrian Jones Good to speak again, Mark. Mark Geoghegan So you work at Hudson structured, but within that group, tell us all about what you're doing? You're you're working on the insurtech side with HS cm ventures, tell us all about it and where that fits in with Hudson and how you're getting involved in that whole insurtech food chain, wait where you play in that in that ecosystem? Adrian Jones So Mark I am happy to tell you about what we do. I have to say it a little bit precisely here. So the description is, we are the venture capital strategy of Hudson structured Capital Management limited, which is a Bermuda based alternative asset manager with about three and a half billion in AUM and committed capital. Here at HSCM Ventures, we seek venture stage equity investments across the global economy that have a nexus with insurance and risk management. So we're looking for opportunities broadly that are connected with insurance. The great thing is insurance is a really broad field. So it's relevant to payments lending, consumer, B2B, SaaS, prop tech, AI, data analytics, cyber, crypto, Health Tech, the gig economy, and numerous other areas of innovation. In terms of who we are, we're five people, we all have some sort of background in the insurance business: life and health, P&C, brokerage, and we're supported by HCM teams in legal accounting, finance, compliance, investor relations, etc. Mark Geoghegan Great, so And obviously, what was your background? Adrian, you were the your previous encounter you were at Scor. So what how would you describe yourself to other insurance folk? Adrian Jones I tell people I was 10 years doing consulting and 10 years in the reinsurance business and then joined Hudson structured two years ago. When I was in consulting, that's how I fell into the business. I was told to go get on a train and go to Princeton, New Jersey and ask for a particular person at a particular office there. And people who know Princeton know what this is. And so that's how I kind of fell into it. I then was head of strategy at RenaissanceRe for six years. I went through the acquisition of Platinum Underwriters with them. Then I went to Paris and as head of strategy and business development at Scor working with Victor Peignet and then John Paul Conoscente and Sylvie van Viet. While I was there, I set up score P&C Ventures that's now led by Will Thorne. With Will, I set that up and ran it through the first fund. And then as I said, I came over here about two years ago to HSCM Ventures Mark Geoghegan When you say venture stage, what do you mean by that? Is that That's not obviously that's beyond seed stage with seed stage is where you're getting a whole lot of college kids who don't really know how to do business plan, any kind of scrubbing, making them presentable. And that kind of thing. This is a bit beyond that this is a business. So you're looking to invest in insurtech businesses that are really they're already happening, right? They're already businesses. Adrian Jones Maybe, maybe not. Everybody gets their first dollar from somewhere. And there are a lot of people that don't have a lot of dollars in their bank account, because they don't come from that sort of money or whatever. And they might seek venture capital very early on, when they just have an idea on a piece of paper, a napkin. Mark Geoghegan So people can bug you with ideas. Obviously, presumably, if they're if they're already insurance, insurance, people that already know how to run insurance businesses, and they got a more technological idea, then they can still they can still tap you up, even if they haven't got a single dollar yet. Adrian Jones There's a couple of functions that we [venture capitalists] can do. We can help to validate business models. We've seen a lot of different businesses. And we have entrepreneurs who approach us and say, “I have this idea, what do you think of the idea, because I'm thinking of, you know, giving up my life to pursue this thing? Is that a good idea or not?” And we'll give feedback on things like that. Other times, people will come to us very early on and say, “I've got an idea, but I need a co founder, I need someone who really knows the tech.” We can try to put two people together in order to launch a business. Some venture capitalists will even help to incubate a business. So they'll say “great, come to our office, and you can have those three desks over there. And you can work with our lawyers to get things set up. And, you know, we will help to make introductions to customers, vendors, etc.” There's a lot of functions that venture capital can potentially provide at the very early stages beyond just $1 or £1. And so when I say venture stage, I'm referring to basically anything which is relevant to venture capital, as opposed to private equity or public markets investing or debt or other forms of investing. Mark Geoghegan And what's your view on incubation? Sounds like you're open to it because from obviously I had Mike Millette on the show earlier in the year. And what what's so amazing about at Hudson, the Hudson grip isn't that you seem to be all the way through the value chain, you could you can do all sorts of things. So again said, If someone comes in, if you're working with someone, does it mean that you know, because you've got, you've got balance sheets, you've got licenses, you've got all sorts of things to those on the table, can you sort of make someone who's appointed representative or that kind of thing? Is that sort of part of part of the package with you guys. Adrian Jones Hudson has several different investment strategies in several different areas of focus, I focus on a particular area. To the extent that we can be helpful across the needs of an entrepreneur, the needs of a company, I can certainly make introductions to my colleagues. When we think about incubation, or thinking about something, generally very early stage, such as helping to form the corporation, helping to develop the product, helping to get the right team around the around in order to develop and sell the product, helping to make introductions with new potential customers. That is what I think of as incubation. Obviously, we can help make introductions to reinsurers, to fronting companies, to other folks like that. But it's generally more on an introduction basis, opening up our Rolodex rather than a formal sort of programme. Mark Geoghegan And obviously, you're helping your investment at that point of if you already have made an investment, obviously, you why wouldn't you help them because if you if you can, you would, there is investment. So there's venture capital investment, which is very passive, and they give some money and put a chequejust trying to get a flavour of that. So it sounds like you're a bit more active than just someone who just gives you money and says get on with him, show me more returns. Adrian From the entrepreneur’s perspective, I think it depends on what they're looking for and what they need. Some entrepreneurs have a really good sense of what they're going to do, they really know the insurance business, and they just want to get on with it. And in that case, then we try to be supportive from the sidelines. And other situations, there are entrepreneurs who have a really good idea, but perhaps it's tangential to insurance or touches insurance or insurance is important to it, but it's only a portion of what they do. And then in that case, we might become insurance experts that they can call as in when needed. And then of course, there are other situations where there's a particular value that we think we can add, and we'll try to work a lot more closely with those entrepreneurs for some period of time. Mark Geoghegan Well, Adrian, you're, you know, you're one of the smartest people I know, and you're one of the people I was go to when I want to get a flavour of what's happening in insurtech well given, particularly given your background that you went through even a little slice of, you know, high end consulting RenRe, Scor. And now Hudson, this is you're talking at the very high end of the sort of, you know, IQ of our industry. What's happening in insurance in the insurtech world? Is it right to say we've been through some kind of a bubble, a bit like all the tech and obviously when we had that massive risk on very low or zero interest rate environment seemed to encourage everything even more put more fuel on the fire. Now we're in this slightly risk of you know, we're seeing layoffs at the really the sort of household name of tech every week now. Where are we at the moment in terms of Insurtech? If we were sort of enthusiasm from investors on a scale of one to 10, where are we? Adrian Jones This word Insurtech is one that we need to reconsider. Because the term was coined in 2015 by people from largely outside the industry, who discovered insurance and discovered that we have made your needs for digitization in this industry, that it's a very large industry -- $7 trillion or whatever Swiss Re says it is today. And so they saw this enormous, enormous market that had a shortage of technically qualified people in it that had enormous needs, and that they thought could be disrupted. And there were a number of businesses which were funded. In retrospect and the clarity now, you probably did see some Bubbliciousness in those fundings. Some of these companies are down by 95 to 98% [in public markets] since their peak valuations. I started writing about a lot of this in 2017, when they first started publishing statutory filings and [I was] saying it looks like there are potentially some issues that people are going to have to work through. The companies are trying to work through those issues, but it's not until after valuation went way up and then came way back down. So what happened, you had a bunch of low interest rates, as you pointed out, that pushed investors to the very extreme edges of the risk spectrum. You had founders and investors who didn't really understand how our industry actually operates. You had Spacs that had misaligned incentives to help take companies public. And of course, just a general euphoria. So the people who are saying now that InsurTech is dead, we're not going to do it, totally failed, just write off the whole sector --those are the same people who only discovered the sector in 2015. MARK: Yep. ADRIAN: The people who have been around for a while, and who have seen it through multiple cycles, including multiple tech cycles, but also multiple insurance cycles -- nobody in insurance is sitting back and saying, “you know, that whole digitization thing, like it just didn't work, like, we were gonna go back to having papers and fax machines and walk around Lime St with, you know, giant bundles in our hands.” Now, there might be a couple of old people, you know, somewhere in the very back of some building on Lime St that are saying that, but the world is moving on. Insurers are continuing their digitization efforts, they recognise the importance of this, they recognise the need to be more efficient in their operations to work down expense ratios, to make their distribution more effective, and to provide better data and information from the start to the finish of the risk transfer transaction chain. So that's not changing. In fact, it's only growing. Therefore, that that's why we need to retire this term InsurTech. InsurTech was a word from 2015 to 2021, maybe 2022. And we just need to retire it and say, this is insurance digitization, or it's insurance, whatever. But the fact of the matter is, it is going on just as much as it ever had. Mark Geoghegan I'm very glad you say that, Adrian, because when I was editing a prominent magazine, we had the idea of thinking, well, we better do an insurtech conference, because so much stuff is happening around that 2015 2016 time. And I was thinking, ‘Well, how long is this going to last for?’ You know, how much should we invest in this? Because how long is it before it just becomes insurance after year five, year six. For me, it surprised me that it went on as long as it did. So you're the first one to call the death of insurer tech, but long live insurance, digitising and just modernising and doing good stuff, by the way, and on the so on that scale back to the scale of one to 10, or zero to 10, or ice cold and boiling hot? Where are we? We sound like we're more kind of Goldilocks, just right. Is that about right? I mean, do you think there's funding available for the right sort of ideas for things that are not going to come out with daft ideas about disrupting everything? Adrian Jones On a one to 10 scale with 10 being hot, I'd say we [as a sector] were at one until about 2014. And I'd say we were at 10 in 2021. And we're probably now somewhere in the six or seven range. So that's pretty healthy. We are observing that good business plans are getting funded. There is right now no shortage of funding for good business plans. It might not be at the terms that the founders want. And it might not be at the terms that people think ought to be provided based on what happened last year. But the funding is there. What's amazing is that there's been this whole ecosystem which has evolved of accelerators, incubators, angel investors, former founders who develop businesses and have since sold them. Venture capitalists, innovation arms and insurance companies -- every insurance company now has an innovation arm of some sort. Startup is no longer a dirty word. And I daresay it was [a dirty word] until about 2016. I don't see that changing. I do think that, you know, perhaps you'll see a natural attrition or change as things go in and out and people reconsider their strategies. But again, nobody's sitting back there saying, “you know, we're going to stop all this digitization stuff, it just doesn't make sense.” Nobody's saying that. Mark Geoghegan What about disruption? Obviously, that was that was an early word that we heard all the time. I remember every conference, someone got up and said, Oh, you know, Uber, the taxi company without any taxis. Airbnb is a hotel company without any hotels, you know, we're going to come up with the insurance company with no capital or whatever, whatever it's going to be. I mean, has that all turned out to be you know, hubris or not. Do you think there are some genuinely disruptive things that have happened? Adrian Jones Insurance is a slow-moving business, but when it does move, it moves in a very big way. And I think the business has moved in the last five to seven years. MARK: Are you going to see businesses like claiming to be insurance but not actually being insurance companies? ADRIAN: Certainly some MGAs have tried to make that case. But the simple fact of the matter is the raw materials of insurance are people, capital, and technology. Those are the three things. There have been enormous advances in technology. There have been tens of thousands of people. It was 55,000 last time I added it up that are working for insurtechs. Capital, you know capital is what it is, but capital is debt, equity, reinsurance. And maybe there's some different flavours of that. But the simple fact is, it's very hard to be in this business in the risk transfer value chain without actually having some sort of reinsurance capital. And obviously, there's a lot of speculation right now about how that's going to potentially be impacting young companies. I think the key for young companies is, you have to be prepared to show that you have high quality underwriting. And that is a big change from what we saw three or four years ago, where, again, the people who came into the industry in 2015, and came out of the industry in 2021, were saying grow, grow, grow, grow, grow, it doesn't matter. The reinsurance will always be there. I don't even know if they thought about it [reinsurance]. But for those people who did and those people who are showing that they can underwrite really effectively, I think the reinsurance will be there, it just again, it will be more expensive, and that'll have to be reflected in rates. But yeah, you have to have capital, this idea of you can create an insurance company without capital, you know, like Uber without any taxis, Airbnb without any hotel rooms, like, I just don't think that model works in insurance. Mark Geoghegan I suppose before, we'll go back to the reinsurance market and how its how it's going to interact, with insurers going forward. But before we do that we should do if we've just declared the death of insurtech, let’s do a little bit more dancing on its on its grave. In terms of an obituary for insurtech, do you think was there anything about it? You know, obviously, we're in this risk of phase with quite sharply increasing interest rates and the cost of capital going up for everybody everywhere? So did insurtechs do anything? Is it just part of this wider trend that's affecting every riskier business and every capital hungry business and startup around the world and whatever sector? Or did we insurtechs do we do something specific? Are there some insurtech specific sort of sins that we've committed? Adrian Jones Everybody has committed some sort of sin. However, everybody has also, I think, pushed the envelope in really important ways to push industries forward throughout the innovation economy. So is insurtech fundamentally different than any other? We created an index called the HSCM Public InsurTech Index [HPIX]. It's an index of 21 publicly traded Insurtech companies. Now, not all of them describe themselves as Insurtech, but these are all companies that went public after 2010, who have novel business models, differentiated by technology and operating the insurance sector. That index, if you compare it to other indices of venture backed companies that went public over similar time periods, has actually performed similarly to businesses outside of insurance. So I don't think that there is much that is specific about insurance, I think that it is a function of a much broader sell off in risk assets that we've seen in the [public] market. Mark Geoghegan It was defensive that way, you still gotta buy insurance ADRIAN: Insurance companies have been around in many cases, since you know, 1850, or 1870, or whatever. They've been through the panic of 1873. They've been through the Spanish-American War, they've been through the disco craze. They've been through at all. And so whatever's happening in the Ukraine, and whatever's happening in other parts of the world, and with interest rates, it's all stuff they've seen before in some form or some fashion. You don't build a great insurance company by changing on a dime because of what was the headline in the Wall Street Journal today. Mark Geoghegan Yeah. Well, let's get back to that, that that reinsurance market because the last time I saw you, I was standing Monte Carlo, I was doing a programme sort of with the whole all the other insurance brokers and all the reinsurance that I could round up and obviously, that was before in the Vedic them was still it's going to be very hard market and one of the hardest markets I've ever encountered. And of course post Ian and it's going to be even harder. Do you think there's any appetite now for reinsurance not affect you saying you know, there's only three forms of capital, you know, equity debt and reinsurance and of course, insurtechs full stack insurtechs have been using leaning quite heavily on reinsurance. There's gonna be any more appetite given the reinsurance market and given the calls on that capital, potentially very profitable business being offered and being declined by reinsurers. Do you think there's any appetite left now foot to bankrupt to continue with someone whose loss ratio is is the wrong side of 100. But with a promise that things may get better, do you say that that patience is likely to run out or they might keep it going for another year or two? Adrian Jones This is a very broad field. But there are a handful of insurtechs, whose rate filings have been analysed fairly extensively, at least in the US. And in many cases, they have been putting up rates faster than incumbents, sometimes by a factor of two to three times. So if the incumbent is raising 10%, the comparison in InsurTech is raising 30%. Insurtechs I have observed taking a very significant amount of rate that has, in some cases quite significantly slowed their top line, which again, is probably the right thing. And they have in many ways gotten ahead of it. I think there are also insurtechs who have very good claim systems who are able to very effectively understand what their exposure is to large loss events, in ways that I believe are actually beating some of the incumbents. I do have some faith that reinsurers recognise that there is value to digitization, there is value to technology. Many [young] companies are fairly well funded. And they’re going to be big reinsurance buyers for a long time. And again, we've all been through cycles, you don't get rid of a good customer, simply because you're in a bad point in the cycle. Mark Geoghegan Okay, so you think some of these, some of these firms are going to come through this eventually, you know, they're going to be having, having tough days now, but you optimistic that we're going to have some successes. Adrian Jones Because there have been successes out of every previous cycle as well. MARK: Yep. ADRIAN: You go back to the 1980s. And there are some pretty notable companies that were formed at that point. In in the 1970s, during the crisis of the 70s, there were very notable companies are formed, then there have been good companies that have emerged from every tech cycle as well as every insurance cycle. Some might not work out, but I think some will emerge, and they're going to be household names in 10 or 20 years. Mark Geoghegan In the UK market, you've got the little red telephone of Direct Line. And that was that technological innovation, that the idea that you could just phone up an insurance company and not have to go to some little broker on your high street or main street, you know, that. That's, that's a household name in consumer insurance in the UK. So yes. And that came from there was a particular revolution around that time, as well. So we should see, we will be seeing some Direct Lines of the future will be beyond their way, you'd say? Adrian Jones Why not? I mean, there's been extraordinary advances in technology across the board. In fact, I added this up recently, I sat down with the Chief Technology Officer of an insurance company and we just drew out all the technologies that you would need in order to build an insurance company, both insurance specific technologies, and then technology outside of insurance, just your core DevOps and cloud and hosting and things like this. And we then analysed when was each of these companies founded. What you realise is you could build essentially an entire insurance company with products that have been launched since 2008 or companies have been formed since 2008, basically, in the last 14 or 15 years. That is not something which was possible 10 or 15 years ago. Even as recently as that, if you wanted to start a young company, you'd have to go out and buy servers and you know, buy hosting and buy a lot of technology or build a lot of technology. Whereas now, if you want payments infrastructure, you just plug into Stripe, and you can have it in a matter of a couple of days. So it is much more efficient to start a business now; it is much faster to start a business. I'm talking particularly on the consumer and small commercial side, but even on the deep specialty side, you look at the number of companies in London for instance that are supporting underwriters who just want to go form an MGA and they want everything else taken care of for them, [you’ll see that] there are a lot of opportunities to do that which didn't exist five or 10 years ago. Mark Geoghegan Yes, I mean yes, if I go to the MGAA This is the UK based MGA trade body, the whole exhibition hall is full of people who are providing plug and play kind of solutions and, and the sale is yet that the underwriter can just turn up, turn up and will will take care of the rest of them, which is taking, you know, a few percent of your business. And that's, that's all it's all about. I suppose. Adrian, what you're looking at is innovation, isn't it? Really. And you've worked, you've worked all over, and you've worked in reinsurance and walked up and down lime street lots of times? I mean, how innovative Do you think my core listenership is? And you know, how, and how innovative is reinsurance? We seem to be always at the, at the back of the food chain. We're backing all these innovators we're supplying with capital, but are we necessarily supplying with much, much expertise and much technological nous? We are always beating ourselves up for not being innovative. But I wonder what's your perspective? Adrian Jones In the London specialty market, the reinsurance markets, the Lloyd’s markets, etc --people love to beat themselves up over, particularly data: “Our data is horrible, and therefore, we can't do any sort of innovation.” There's been $50 billion which has gone into insurance innovation from venture capital, and about 40% of the deals that have been done, [measured] by deal count, have had some insurance strategic investment in them. So insurers and reinsurers have absolutely been backing young companies. And about two thirds of the money which has gone into the insurance technology sector has gone into carriers and MGAs. Ccarriers and mgas, of course, all have to have some sort of either insurance or reinsurance component to them. So I think the market has actually stepped up in a massive way, in terms of supporting innovators and being willing to take the risk that the innovation goes wrong. I think that's something that we should all be very proud of, and very happy for. Unfortunately, that doesn't take care of the day-to-day things like needing to have wet signatures and issues like this, some of which is driven by regulation. And I think many regulators are trying just as hard as everybody else to get rid of some of those sort of sticking points and snags. Obviously, in the London market, there have been several different attempts at digitization, a number of which have not gotten particularly well. But I'm actually pretty optimistic that, again, to the point I was making earlier, the technology that many young companies are using didn't exist 10 or 15 years ago. And so I don't think that we can take what happened 10 or 15 years ago and say, “well, therefore it's going to fail now”, because the technology has massively improved. What has not improved, where there has been zero progress, is around federating insurers, brokers, reinsurers and other interested parties in the ecosystem in order to drive towards common goals. And we saw this with B3i, for example. People were very happy to use the bridge once it was built, but nobody wanted to pay to build the bridge. Mark Geoghegan This is the industry wide blockchain initiative, which has now I think, since folded, hasn't it? It was it was sort of set up with big fanfare about three or four years ago. Adrian Jones The industry put a lot of money into it, and a lot of effort. But there was always a free rider problem. There were always people have said,“why should I pay to build this? And by the way, is it going to be built the way that I want it? Or is it going to be built the way that somebody else wants it?” There was a lot of parochialism, unfortunately, that that was observed in the industry. And this is just among the reinsurers and the carriers This is without even bringing in the brokers. And that's a whole other ball of yarn. So unfortunately, the industry you know, we do come together from time to time and we make efforts but I think that the industry has to be more willing to set aside parochial interests in order to promote the common interest, which is to deliver better insurance at a lower cost to consumers. And everybody talks about that, but I think that's an area where the criticism is quite deserved. Mark Geoghegan But in generally you did say at the beginning that you bit more optimistic than you have been. Adrian Jones Yeah, Mark Geoghegan Yeah, things are happening. So I bet you know, you encouraged by some of the things you see coming out of London. Core data standards and some of the real basic building blocks of what we're going to need, if we're going to digitise this insurance world. Adrian Jones I'm excited about what Lloyd’s is doing – the core data record, Blueprint 2.0. Obviously, these are long, complex sort of projects. But I think that's exciting. I think the Lloyds Lab is very exciting. We've seen a couple of pretty interesting companies that have emerged from that, that are doing great things around the world. So yeah, so there is there is definitely progress. Mark Geoghegan And how far off do you think this fully digitised, completely digitised insurance value chain, you know, going all the way through from, you know from Main Street all the way to retro and the sort of things that might get involved in, you know, the ILS and the capital markets, you know, where some little bit of data can be flowing all the way through into an insurance company, into a treaty into something else, all the way to retro maybe? Adrian Jones There are two challenges that our industry faces before that's going to happen. One is, we are extremely heterogenous, in terms of the types of business that we have, the way that business is processed, the way it's underwritten, the systems that are used, etc. This starts all the way back at the brokers. I was in a conference with a bunch of independent agents earlier this year, and somebody got up and they're sharing best practices. And they said, “the industry classification codes that we use in our brokerage are just horrible. And I thought, how are we going to get this right - carriers were not happy with us.” And he said, “you know, the solution that I came up with was, I trained our receptionist to look up the codes. And our receptionist is really good at doing that.” And I just I was shaking my head because this is where it starts --it starts at the agency. So garbage in, garbage out. But number two, there are great tech solutions, which can do that [finding industrial codes] very quickly and easily. When the attitude is the industry is “let me train my receptionist to do this”, versus “let me find some great tech that will do this: -- that sort of attitude is what really has to change before we're going to start seeing real change in a business which is as heterogenous as the insurance business is. Mark Geoghegan Is it just a cultural differences that just someone who's 20 Years Younger would think tech first rather than think of just getting the receptionist to do it! Adrian Jones Well, the average age of an agency principal in the United States is 55. Yeah. Mark Geoghegan Yeah. Adrian Jones And they're having a hard time getting young people to come into the industry. Mark Geoghegan It's probably not surprising if the first five years of their job is classifying risks by some mind-numbingly boring manual method of inserting a lot of codes! Adrian Jones Then you're going to put those into an agency management system, which probably hasn't been updated in quite some time, which is a relational database. It's not particularly cognitive. It feeds into some relatively old technology for how the agency and the carrier interact. And then, of course, some carriers still have mainframes. So there are challenges all along that value chain. That's the second issue, which is we have a complex value chain, which in many places is intermediated. That is different from the bank value chain, for example, where a bank can own the customer relationship, basically, from start to finish. And so banks can be much more effective at pushing digitization because they control the whole thing. Whereas because reinsurers have outsourced their sales force to brokers and primary carriers have outsourced their sales force to independent agents. It's much harder, because you have competing interests, and you have people who are going to want to do their own thing, and you really can't tell them what they have to do. Mark Geoghegan Particularly when they start to think that their way of doing it is some kind of gives them some kind of competitive advantage. And when usually it doesn't, then that's the real problem, isn't it? Adrian Jones It's only a competitive advantage in a very narrow sense. Yes, it's not a competitive advantage, which actually benefits the consumer in any way. Mark Geoghegan Yeah, absolutely. Obviously, I spend most of my life looking at the wholesale specialty and reinsurance market as you know. Where risk is complex, it's probably large, or it's just weird, and it probably crosses a border, even if that's just crossing a state line from one place to another, you know, in the crossing into the E&S market, the very least in that market Obviously, we always like to think that you know, it we're we've higher up the food chain, we're a bit cleverer, we're more bespoke. Everything we do is complicated or weird or difficult and hard to understand or harder to aggregate. And there's less data because there's probably less number of risks, you know, then the list of those risks are not so homogeneous. That's often our excuse for not automating what we can automate. How far do you think so a lot of people that I talked to, we've got a business like Ki syndicate in Lloyds just been given a big preemption actually approved a 50% preemption up to 1.2 billion. But that really is the exception at the moment, rather than the rule. Most of the people that I talked to are saying, well, no, I'd rather lead business, I would like to turn my underwriters into more sort of bionic underwriters, I don't want to be algorithmic underwriters. I want my specialist knowledge, you know that these are guys who know about fish farms and helicopters and whatever else, and war risk and whatever, a huge amount of specialist knowledge but, and I'm going to use the tech to kind of, perhaps triage their day, send them only the things that we think we've got a good chance of winnings that we don't waste their time, with, then I'll score the preschool the risk, I'll look at some of the things I'll bring up all of the back data and the satellite imagery and whatever else they might need to do their job. But I'm still gonna leave them to do that job manually at the end, but I'm going to enhance them digitally, rather than take them over by an algorithm. Do you think that's the right way to do it? Or do you think more things will be automated? And do you think there'll be more algorithmic underwriting as well? And because that algorithm is not quote and bind It's follow yes or no. And what line size Adrian Jones When I came into the industry 10 to 15 years ago, I saw that the leader of the slip gets paid essentially the same as the follower on the slip. And the leader does a bunch of work in order to understand what's actually in business. And the follower just puts down a stamp, and yet they get paid the same. I said, “wait a minute, the value that's being added to this process is very different.” So why can't you have both? Shouldn't the leader have access to as much information as they possibly can about how to underwrite the risk, and perhaps even a computer which is going to recommend a particular price? But then you have the human element that gets added on to it to say, “Does this really make sense? Is this the way that I would do it, et cetera?” I think that's absolutely valid. And once someone has done that, or perhaps once two or three people have done it in order to put together lead capacity, if that's what's required. Why should anybody else look at it? Mark Geoghegan Shouldn’t maybe have an algorithm looking at it, just see if it makes sense, from a financial point of view that there hasn't been a mistake? And you know, fat finger errors? Maybe you've got a zero in the wrong place? That kind of stuff? I mean, you don't want to have that we see those kind of horror stories on trading floors every other day, don't we? We don't want that happening in the room in Lloyds, do we? Adrian Jones For sure. And there are people who say, well, algorithmic trading has never actually worked. And so it's not going to work this time. I reflect on that and I say, yeah, there are very valid reasons why it hasn't worked. It's very difficult to get right. It takes a lot of effort. But I come back to the point that I made earlier around the incredible advances in technology just within the last 10 years. And I think people underestimate that. I'll give you just one statistic, the power of AI classification models in the 10 years from 2010 to 2020, improved by a factor of 44x. MARK: Wow. So these are - totally rubbish to being quite rubbish. Adrian Jones I refer to classification models that do things like telling: is this a picture of a dog or a wolf? And back in the day, they would have said, “that's a picture of a wolf, because there's snow in the background.” They've gotten a lot lot lot more powerful, a lot quicker to run a lot simpler to run, and the technology to run them has greatly improved. This is a little bit like the situation that we had back in 1993, when cat modelling first came out. Cat modelling was very much pushing the edges of what was possible from a technology and computing perspective. And there were certainly people who resisted it. And there were certainly issues and still are issues with cat models. But the simple fact is, the people who had very good cat models ended up doing very well. And you could say, well, that's an algorithm that just you know, that just tells you like that the expected loss is going to be and therefore, you should price it at this rate online, whatever. I think that this is just a natural evolution of some of the trends we've already seen in the space. Mark Geoghegan Because obviously, models may not be perfect, but they're a hell of a lot better than they used to be 30 years ago, Adrian Jones and they're a heck of a lot better than somebody who puts pins on a map of where my risk is and relying on payback. Hurricane Andrew blew that up because it blew up half the sector. So we have to be much more precise nowadays. Mark Geoghegan And so do you think we've ever seen an algorithmic lead in something quite complicated or something quite, you know, specialty. Obviously, we've got algorithmic follow that seems to be doing perfectly well. Adrian Jones I think that a lot of underwriters have algorithms which produce recommendations to them, and sometimes even pricing recommendations. I think, actually, that's been the case for some time. They just don't talk about it. Because you don't want to go to the broker and say, “this is the price because that's what the algorithm said.” Mark Geoghegan Yeah, because otherwise, we'll just reverse engineer the algorithm and then start gaming it. Adrian Jones Well, that's kind of what happened with cat models, right? People just all started turning down the risk all along the way. And using the cat model which happened to give them the more attractive view of risk, and we see how that ended up. So there always has to be a human. The question is, What is the incentive of that human? And can that human maintain pricing discipline? And in times when they can't, what do you do? Mark Geoghegan 11:06 Certainly, yes, of course, whilst we might have the perfect algorithm, we still don't always have the perfect insurance market. And the temptation for the manager of the algorithm will be to start dialing things down at some point, because otherwise, the thing is just gonna write zero business, isn't it? I mean, isn't it always the human element? Some of these things, I remember, when I was a young broker, my boss told me I had a risk was declined to me, and I was a bit upset. And he said, don't worry about it, you know, this person is an underwriter - underwriters have to underwrite. At some point, someone will underwrite it, don't worry. And this kind of was a lesson I learned. And I do think that's the case, we're still we're always going to have that human factor because of, you know, people want income and they won't want to turn all the machines off. Adrian Jones We're always gonna go through up and down cycles, where there's always going to be people involved in some way in the process, and those people will have their strengths and their flaws. So yeah, I mean, you have to be somewhat commercial. W here this can go wrong is again, I think people get tempted to start changing the algorithm to adjust down the view of risk in ways that don't make sense. And this is where governance of algorithms becomes really important. And I think an area that regulators are starting to focus on more and more. How do we make sure that people aren't adjusting the algorithm in ways that turn down the view of risk that ultimately blow up? And there's not a good answer to that question right now? Mark Geoghegan So we need sort of an algorithm to police the algorithm and police a person doing playing around with the algorithm and then yes, and just Yes, bending the curve to fit to fit their particular appetite to fit the market, rather than making trying to make the market bend to fit the actual curve, Adrian Jones Or even just inadvertent drift. You can set up a model and make it perfect, but machine learning models learn over time. And so do we understand how the model is learning? And what sort of changes are resulting from that, and are our regulators, or Lloyds or whoever, effectively able to understand how these things are changing? I don't know. I think this is a pretty emerging area., Mark Geoghegan they might learn to be greedy and fearful, and both at the wrong times. Who knows? Who knows? Let's get back to insurer tech, what are we going to have to call in shorter, we're going to call insurance, innovation and digitization. We’ll have to call it something else in which you operate in the venture world of insurance in funding new companies that do clever things with insurance. Adrian Jones I mean, I guess we need a name for it. But, I've always hated the term innovation, because the first week that I was in this business, somebody sat down with me and they said, “I want you to understand, you shouldn't use the word innovation. That's like the I word. It's like a bad word. Because there have been a lot of people who don't really know the business --and by the way, Jones, you don't know the business yet--who have come through here and who have talked about innovation. And you know, it just hasn't worked. So don't talk about innovation, talk about like how we're going to make the core business run better. That was the advice I was given. And I think that was absolutely very sage advice. And it's also advice that I pass on to others, like, don't come in and try to like blow things up or change them overnight. This is a business which does move and when it moves, it moves in a real way. But it's not going to move overnight. Mark Geoghegan If you can come with a really great proposition really well tested to practically prove to me beyond all reasonable doubt that if you do this will reduce my loss ratio, my core loss ratio by three points. Then come and have a conversation but if not, maybe you know I'm not that interested. Adrian Jones But let's recognise you can't actually prove that. How do you how do you prove that something is going to reduce the loss ratio? That's impossible. Mark Geoghegan Well, yes, because you just don't know. I mean, losses are random events as well. I mean, let's face it. We might just be really unlucky. Adrian Jones Your risk is constantly evolving in its nature. And so something that may have reduced last year share by three points in a back test might increase it by three points in a going forward test. And there's always the commercial element to it as well, you can't always replicate that exact same portfolio, because if you change the price, you're going to get a different type of customer that comes to you. And those customers might have different characteristics that you don't know about. Mark Geoghegan We’re in this slightly cooler, let's say it's it a more rational investment environment. Does it make it easier for you to operate? Sure, hope so. You don't have to compete with people who have got unrealistic expectations, or people who have been overexcited, or, you know, willing to support what, with hindsight, were very frothy valuations, or people who've discovered as you said earlier, the sort of people who only discovered insurance in 2015, and now have now sort of knocked off the Christmas card list. Adrian Jones If the world was filled with crusty old people who sit in the back of Leadenhall Market with a beer in their hand, and say, “That's never gonna work, I've seen that before, it's not going to happen, don't even bother just, you know, keep doing things, we've always done it” -- we would never advance. You need a little bit of that. And you have to respect that viewpoint and understand it. But at the same time, I think the industry needs people who are a little bit crazy, you come in and say, like, “gosh, you guys just don't get it, here's a better way to do it.” And the history of the London market actually shows that there are a lot of people who have come along -- we can all name some great entrepreneurs there -- who have come along and said, I've got better ways to do this. And in many cases, they've leveraged technology in some form, but not always. I think there's absolutely room for all of these people in the sector. The challenge is keeping things in a little bit of a state of balance, and not getting too over enthusiastic with people who are going to spend a lot of money to learn less than this the hard way. Mark Geoghegan And what are the common themes that are getting backing today? What is there anything that successful insurtech entrepreneurs have got in common that is that is getting them positive feedback from investors? Adrian Jones My observation is, we are still seeing good companies getting funded. So what is good -- it's a lot more about the characteristics of the people who are involved. We are seeing a much greater emphasis in the market on people who have a little bit of grey hair, who have some insurance expertise who've been around the block, who can say, “I've seen this business for the last 10, 15, 20 years. And I know that there is a better way. And here's what it is.” Often they are partnered with a technology person, or multiple technology people who say “I can build a technology in order to do that. And I can build a modern, efficient, effective tech stack.” They often come with relationships, if they're going to do it as a carrier and MGA, they're often coming with relationships into those markets. So they have people who are they're known quantities, they're willing to back them, et cetera. Outside of the carrier/MGA space in SaaS, data analytics, etc., we're seeing some companies that have actually been around for a little while, where perhaps they were not the hottest thing back in 2020 or 2021, but who plugged along and actually have built really interesting technology, We're starting to see that those sort of companies are being successful in getting funded. The challenge right now is for companies that got a little bit too caught up in the Bubbliciousness, who perhaps raised a little too much money at too high of a valuation and spend that money but don't have what they need to prove that they are qualified for the next round of funding. We haven't seen a lot of failures yet. We've seen a few. Mark Geoghegan Yep. Is there any particular part of the value chain that's hotter than any other part at the moment? Obviously, you know, looking from my perspective, I've seen so many, hybrid carriers, but I wouldn't describe any of them as insurtechs necessarily, so anywhere within this insurance innovation, and now I can't say innovation. I can't say insurtech, so I don't know what I'm going to say I would have to say insurtech for the moment. Are they playing in a particular area, are they congregating in any particular part of the value chain? Adrian Jones Here in the States, we have seen very few direct to consumer companies getting funded over the last several years. Much more emphasis on specialty, and specialty in a very, very broad definition. So, that can mean specialty classes of risk. It can also mean specialty services into insurance. It can mean specialty finance around insurance. People say “in niches there are riches.” And I think that's true in the insurance business as well. So that in order to do these sort of specialised businesses, that does require that sort of grey hair skill, and partnering between, you know, old people and young people, people who are experienced people who are inexperienced, etc. That's where we're starting to see things. But again, it's much more about the type of the type of people who are involved than it is, I think the sector because there's so many sectors and insurance that it's really hard to say [something like] people are not focusing on P&C anymore, they're only doing life and health. Like that's just not true. There's opportunity everywhere. Mark Geoghegan Another one of these buzzwords of the last 12 months, and I'm sure one that you will want to retire at some point in the future is embedded insurance, it seems to be a thing that everyone wants to talk about. And I don't really understand what people are talking about when they talk about it. Is that something that that we should all be taking more note of? What do we really mean by embedded insurance? Adrian Jones There's a million definitions of embedded insurance. I think of embedded insurance, basically, as being not insurance which is sold via your traditional independent or tied agent channels, and which is not sold via direct response advertising, whether that's an 800 Number, a red telephone or an ad on Google, it's not insurance sold by price comparison websites, and it's not insurance, it's sold to some sort of lead gen. What it may be is insurance sold through affinity. And of course, affinity is just about as old as insurance. You know, the people in Lloyd's coffee shop, they were probably an affinity group. It’s point of sale, so involved in another product’s point of sale. It's some other form of non traditional distribution such as distributed via retailer via vertical software company. I put bancassurance in as well, you know, insurance sold through banks. And I would put in insurance which is sold via some sort of non-insurance transaction. So it's kind of a woolly concept, but it encompasses a lot of those different things. Mark Geoghegan Doesn't sound like anything new to me. I mean, these are all as old as everyone knows that you buy a holiday, you got to even the old fashioned travel agent that you walked in, and they had little computer. They were still trying to sell you travel insurance. Of course, they still do that now with the budget airline. Now you did on the website, there's always the page with the hire car, the hotel, and there's always the travel insurance as well. I mean, is that embedded insurance? And what's really special about that? Adrian Jones By many standards, that would be considered embedded insurance. But travel, travel is one area. And I think travel is what people think of when they think of embedded insurance first, because it's the most tangible and in many ways, the most obvious. But take another one, some of the largest sellers of small business workers comp insurance are payroll processing companies MARK: That's interesting. Why? ADRIAN: Because they have the data, they have the data on who's getting paid what. I mean, it's all rated on payroll. So they know what the job descriptions are, they know how much the people are getting paid. Mark Geoghegan if it's priced by a big algorithm, then of course, they can do it. Adrian Jones Yeah. So they're selling workers comp insurance alongside payroll services, because they have better data on what the product is, what the person is, et cetera. There's a vertical software company in the US that does construction management software. So they, you know, they help a general contractor and all the subcontractors to stay coordinated and know who's doing what when, what's the plan, they manage, they record safety incidents and things like that. And they set up an agency where they're starting to sell insurance . Mark Geoghegan It's going to be more sophisticated, but you're saying some of these people, of course, probably know more about risk than we insurers do. And so it's the sort of thing where we might you might get some massive Siemens turbine that costs $50 million for a power station and, you know, they might embed insurance in and that is that because they understand everything about it. Adrian Jones In some ways, that's just a product guarantee or product warranty -- guarantee the turbine is going to produce a certain amount of power. And if it doesn't, then, you know, we're on the hook for it. Is that even insurance? In many cases? It's not. That's just the seller adjusting the price based on performance. Mark Geoghegan I suppose it could interact with insurance because insurance can get involved that's when a sophisticated amount of data being shared, then an insurer could help. And then that's where insurance does good things, isn't it? Adrian Jones Yeah. Because here's the challenge. The big challenge is that data is friction. So everybody talks about data and data is the new oil. I think that's another statement that needs to be retired because, number one. Oil is finite and data, in many ways is infinite. But, number two, oil is a lubricant. Data is not a lubricant Data is a friction, at least at first. The person or company seeking insurance has to go and produce a bunch of data, then somebody has to take it in and analyse it, and reformat it and do whatever they're going to do with it, run it through an algorithm or run it through a human's brain and spit out a price. That's a lot more frictional than just saying, "well, you know, last time it renewed at this price, and we're adjusting all the prices by 10%. And therefore, it's gonna go up by 10%.” That is a much smoother transaction, and in many ways is actually much easier to accomplish. But is that really the right thing for the customer? Is it even the right thing for the industry? So the challenge around embedded is how you turn data from a friction into a lubricant. And I think there are people who are doing some pretty interesting things there, but it's going to take some time. Mark Geoghegan Goodness, yes. No, I didn't even know how we could begin to, we'd have to do a whole new podcasts on that one. Adrian. Thanks so much for your time. I've come to the end of most of my questions. I think I was just having a quick scan of what I plan to ask you. I think I've asked everything, and probably quite a lot more. So. Thank you so much for your time, Adrian. It's fascinating what you're getting up to and I hope you'll come back on the show and give us an update at some point in the future. And obviously do tell us when you worked out a new word for insurtech. Insurance technology, Is that that we settled on? Adrian Jones It's good enough for now. Mark Geoghegan Excellent. Thank you so much, Adrian. Adrian Jones Thanks, Mark.

0 comments

Recent Posts

See All

Kommentare


bottom of page