05 Dec 2019
RSA bounce short-lived as markets decline again
Insurance Day Business Briefing: London
A panel of insurance industry experts discuss the product development and innovation in insurance.
Don't have time to watch? See the full transcript below.
I would like to move on to more about developing new products and product innovation and how the new technologies that you’re looking at can help, new data can help develop new product lines. What’s your perspective on that Vincent?
I mean the best example would be the partnership with Oxbotica. So, Oxbotica are a specialist company that builds software or the brain that fits in a driverless car. What’s interesting about their technology is that it can be applied into any mobile system. So, for example, you could apply that technology into a fork lift truck and then convert it into an autonomous and it can run round a warehouse. Or alternatively, it can be applied in a driverless car; alternatively, it can be applied in a heavy construction vehicle. But what’s really important about that is that is bringing a completely different shape to the risks and the profile that where a lot of the human errors relating to that, so 90% of traffic accidents are related to human error. But then that’s translating to a very different nature of risk. So, clearly cyber is going to be a very, very, very key risk.
So, interesting when you look at autonomy, one of the big differentiators is the speed with which these machines learn, is much more exponential and the reason is that the machines share their learning and therefore, you get much more rapid development around the experiences that they have. But that means they’re more connected. Which means they’re much more exposed to connectivity risks and cyber risks. And similar to around product liability for obvious reasons.
Our work is around understanding what good or bad autonomy in working with our partners, but then thinking about, well how can insurance propositions be different in those contexts. And so, it is actually working with the partners to define that and shape that and really starting to reimagine what insurance may look like.
So, the claims process, it would be really mad to have the same claims process that we currently have when actually all the telemetry in the car is already telling you what’s happening and when it’s happening and how would you recreate that process to have something feel and look completely different and that’s what we’re doing.
And what insurance problem did the work you were doing solve? If you are trying to understand the risk associated with an autonomous vehicle, what did the work that you are doing look to address, to get a better understanding of the risk? Is that what you were trying to do?
Ultimately, the first is about the risk piece but there something specific to driverless cars and that particular sector. But, specifically, to autonomy underpins huge ranges of the sectors involved and the impacts on different industries and sectors and we feel, as a specialist insurer, we feel that this is really, really important that we understand that, so we can help our clients understand what good or bad autonomy looks like.
You can expand that understanding to other areas, not just vehicles, but other areas where there is an element of autonomy, such as, shipping for instance?
We’re already doing that. I mean warehousing would be a really, really good example of that. So, the technology that’s being used within Oxbotica, you can take specific aspects of that technology and use that in different environments.
Okay. James, coming back to the general question about developing new products, how does the innovation work that you’re doing, whether it’s finding new technology, or whatever, how do you actually turn that into a new product?
I think the starting point and again, a lot of common themes across all of this, is client need. So, if you look, if you go back to the list of unmet client needs at the moment, so, I’ll just go to the top ten and for say manufacturing, or healthcare, whatever, they’re all pretty similar, right? IP protection, massive issue, hard to cover; you go to things like business interruption, still a major issue, hard to cover; brand reputation, still and issue; regulatory risk, right? These are all interesting areas of real client need, in their top 10 every single time. Very few solutions. Our starting point are those, our starting point are the client needs.
Then you say, “Okay, what sort of solution can we build around those?” And what we’re finding is, we are starting to have to work with partners to get the information and the data that’s actually going to help us create some form of protection. And if you take something like IP, you need real specialist people to understand the IP. So, we brought a new business last year, who are a cyber specialist firm, they happen to have a group of people who are really deep into IP and understanding and the value of IP. We have been able to then take that group of people, and actually link them back to the market here and start to get real data and information that can give markets confidence to start writing protection in the IP space.
So, we have to find the need; we then have to find, often now, new data or new information and bring that back, with the need, back to the market to work with them, to create a new solution that will meet the client need. I think what’s interesting here and the bit of innovation for me is that, if you bring it back to London, people are having to accept different forms of data. In cyber, no actuarial tables, no long history. This thinking of: how do we accept something else and still create a market? And how do we do that quickly? Because, again, cyber’s an easy one to pick on right at the moment. Everybody sort of likes it; there’s probably one market that’s taken a pretty big bet but broadly, there’s still a huge amount of conservatism. And that’s the issue. So, we’ve got ways of building new products.
In my prior job at Aon, I worked with many, many different insurance carriers across the globe, fantastically interesting. And a lot of them said, “We want to innovate, we want to write more risks.” So, we did this really interesting analysis, where we took our entire portfolio and we broke it into 18,000 different segments, by industry, by product line, by country, by client size and then we looked at the level of competition in each of those segments. And what you find is, and somebody said it earlier, it might have been Paul earlier, saying, “People like portfolios of low volatile, commoditised, risk areas. I can tell you, in all those spaces there are 18 competitors, maybe 20. It’s just price; pure price. You reach out and you go into the areas which are harder to understand and you need new ways and a more volatile. You go into US trucking, I can tell you, there’s not 18 competitors in US trucking.
So, what’s really interesting for me and the real opportunity here around product innovation is, and I am happy to talk to anybody about where the opportunities are because our clients need protection, but what is it that you need as an underwriter to be confident of actually protecting that risk. And if we can unlock that as a market place, then actually, there is huge growth because the irony is, I go back to my 3% - 1.7%, there’s plenty of risk out there, we just haven’t quite solved that new data, real client need, bringing those two things together but there’s big opportunity.
So, is it about finding the data to back it up or is it probably a combination of getting the underwriter to feel comfortable using that data?
Well, this is the irony, right. It is about finding the data, and then guess what, it’s about taking risk. And I would say again, Solvency II, some what? And other factors right, have reduced down our appetite of risk dramatically. In this world of… Our clients face massive different issues on them all the time. Changing environment, they work really, really quickly. We all do in one way, so their needs are changing but we as an industry, have worked in about a hundred years of historic data and a line that runs through it and then I feel comfortable. How do you turn that into, “Hey, two years would be fine”? And I think that’s the challenge.
I don’t know whether you’re the right person to answer this but how does XL Catlin approach being presented with new data sources and acceptance of new data sources which may not have the 300-year history that perhaps some classes did?
Certainly, the lack of claims experience is critical for those sectors but again, that comes back to Mike’s point about relevance is we do need to find opportunities and propositions and how we deal with that. And the way we approach that and the way we think about structuring that may be very different from the way that we looked at that in the past.
And again, thinking about pooling and thinking of other arrangements, it’s got to be, there has to be a different way.
But I wanted to just come back to the data part. Actually, thinking about how we can pool external data, and use external data, is really at the heart of being able to get that comfort and get that understanding. And so, a lot of the work that we do is around how do we source that data. So, Cytora, the company that we partner with, has enabled us to go into new markets where we haven’t had claims experience before. And what they’ve been able to do is actually help us understand the insurable interest within that market, help us understand those clients, help us understand their digital footprint, but interestingly, try and recreate the market loss history for that sector.
Can you give us some examples?
I can’t give a specific example because we haven’t yet launched it. But there is the ability to do that. It isn’t perfect because there is some judgements used in needing to fuse data but it is certainly much, much, better relative to not having any experience. And it does come back to taking a risk and learning, and fine tuning that as you go along.
Can I bring you in here, EXIN doesn’t have a legacy approach to looking at risk?
Not yet. But that is definitely something from the beginning that we thought about, which is how are we going to get that out. So, on top of creating those paths, or whatever, which are going to get the data, that’s what I said to my team, “You have to be data searchers. So, you discover if you search well, you discover huge masses of information, coming from the government, for example. You have a big effort all about the world of governments to put some data on the open market, I would say, so there is tremendous information there, especially all the policy reports for some big cities and so on and so on, which you can collect and use. And this is for free as far as you can declare this information comes from there, you can use it.”
So that is fantastic and then if you search the web, the way that Cytora does it, which is looking at, if I’m searching for, is it a good business to insure the butcher around the corner? Then searching about the whole stories, you can find about butchers having issues, or people complaining about their quality, or whatever, everything is there. You just have to collect it and know how to trade it, to remove the noise and get the information that is really valuable for you as an insurer. Am I taking a large risk or not insuring that one and if it’s a risk how much should I price it? But everything is there. So, it’s just finding that information, so you need to have people who know how to search and look for that information but it’s there.
Do you feel that perhaps EXIN has a more open appetite in its willingness to accept risk, whether it’s perhaps a shorter history of data than perhaps a longer standing insurer that may have, perhaps, a different approach?
Well usually we don’t have the 300 years of history, so from the very beginning we are going to start with the data that we can get, even if we are going to have some historical data from some acquisitions that we are working on. We are going to mix those, to mix the fresh data that we can collect with the historical one that we have and also potentially why not, partnering with some companies that have data, which could help us. So, in both ways they are going to provide us with data and some of the insights we are going to get, we are going to give them back those insights to help them in some other part of the business; but that is also an exchange of the data.
So, from the very beginning we thought about that way being a hub, where we could grab some data and exchange this with some other providers to make sure we achieve what we want to have in the end, which is a good understanding of the risks that we are going to take. And yes, we might be going on some risk areas where nobody else is there, because from the very beginning we understood how to price it and understand it.
Do you think the insurance industry can learn from other sectors in terms of how it approaches data?
For sure, I mean look at what happened in the entertainment industry and the way you consume music and so on, and how the people are pushing you. Now music which is exactly matching what you love and what you’ve been listening in the last ten days or whatever, or matching your mood. I mean that’s just crazy because the way you move and your mood, you could be pushed some music which is exactly matching the way you are behaving that way. So, yes, you have to learn from that and it’s going to massively change the way we create our products and distribute our products, that for sure.
Just on that, it always interests me that we sort of to some extent as an industry, and I’m talking broadly now, look at ourselves as not having been innovative. I think the UK insurance industry, and I’m not talking the London Market, but the UK Personal lines market, particularly around motor had been an unbelievably innovative market place. More innovation has come out of Cardiff than Silicone Valley in insurance, in the last 20 years.
And it’s completely under-rated. If you look at how it changed from, I don’t know how many years ago, but let’s say 25 years ago, then along comes a player like Direct Line and starts using the phone and going direct, completely revolutionizes it and then guess what? It gets completely massacred for a while and they’ve done a good job of coming back, by these start-ups in this industry, with really interesting capital models, really interesting operations, really interesting, deep understanding of clients and pricing. And players like Hastings and Admiral before Hastings were able to compete in what is, probably, the most transparently, competitive market anywhere, in a highly price-driven environment. And do that and make money, and do that and make money without massive scale. I mean it is remarkable; this is a group of highly entrepreneurial people in insurance, in the UK today. And yet we sort of feel that other industries have been more innovative. I don’t think London’s been, London was highly innovative at times in its past; I’d say now it struggles with the speed of innovation at some level but as I say, insurance itself, remarkable, I think in some ways.
How can London speed up, what can it do?
Well this is the problem, so I go back all the way to my book at the beginning, the issue of being a large organisation trying to innovate is that you have, the fact that you are a large organisation means that you have a pretty attractive business model. You measure yourself by your return, or your margin, or whatever route you want to do. The irony is that the disruptive model, that you were saying earlier about destroying your business, when you start off, that new model looks pretty unattractive. You’ve got to invest, the return’s not great because there is no scale in that model today; so, where are you going to put your resources? You are going to put them into your current model because it looks better from a financial return, you’ve got a CFO, who’s been on your back and you can’t possibly…
So, you have this inherent why entrants can come in and these other players can come in and innovate because you can’t. So, I think what can you learn, well actually Direct Line had this long journey and, as I say, they’ve come back pretty well but they, for a while, got pretty heavily beaten and it’s a good lesson. You can’t, if you really want to be here in 20 years’ time, you can’t avoid accepting that you’ve got to do some things differently and you can’t avoid that that will have a lower return than your current business. If you don’t accept that, I think it’s extremely difficult.
Part of that comes to senior management and senior management saying, “We are happy to accept a slightly lower return on business because we are going to take a long-term view.”
Or you have to act as a software, or life science business, which is, “I’m going to invest a couple of millions now for some products that are going to bring return for the next 10 years, starting in two years from now for the next 10 years.” Which is a completely different mindset than a pure service business, where you have a need, okay, you hire a couple of guys, you match that need and you do it; like the consulting business or whatever. There you have to think, “What I am investing in for the next two years, this is what is going to help me make money for the next 10 years”, which is a completely different mindset from a financial perspective and from the management of your teams, definitely. You have to accept that for two years those guys are not going to produce anything but if you’re not doing that, in two years from now, when they will be ready, if you have not done that, somebody else will have done it and you are there. They have taken the market.
How does XL Catlin approach this?
Yes, I think there is a really interesting aspect around. So, what we are actively doing, we’ve got a broad portfolio of experiments, where we want to be out trying, doing very specific things and very narrow tests. But each one of those ideas can be scaled across the organisation. So, we always say, the ideas, we want to think big but we want to test really narrow and really focused. So hence if those things do fall over, and some of those things do fail, and if they don’t fail then we are clearly not innovating, that we need to be able to then work out and be confident that we can create a patchwork of ideas. I think someone talked about it earlier, was around, be confident that actually through those experiments then the new models will emerge from those patchwork of ideas. But you won’t have all of that picture right at the outset.
The really important point for the organisation is when do you scale those ideas? And that’s when through those experiments where you build the confidence and you build the buy in, working with the business and the business leading those ideas, of saying there is something really, really, genuine here, that is really material for the organisation.
The last part, which is then we clearly need very different skills, very different talent, complete cross-pollination around that but more importantly it’s a really different mindset and I strongly, one of the hardest things is around this idea of sharing, what I call, half- baked ideas. So, to really innovate, we need to rapid prototype and share those ideas and share with different people but actually it’s really, really, hard, you need to be quite brave in sharing ideas across an organisation that are half-baked. It is through the engagement with others is really how those get shaped. I think that cultural part is really important and how we start to recognise that but actually embrace that.
Just one thing, the way I’m pushing that in EXIN is to say there is a rush, we have to go fast but we don’t have an internal need to be the first. We can be the second one or third one in the market but we need to be the best. So that’s why we need to share, learn from others, not necessarily be the first in the market but understand. Because being the first in the market, you have to evangelise and I can tell you this is want happened to me 15 years ago when I came on the market with some ideas that algorithms could change the way we trade on the stock market, or change the way you design your product, or you can change completely the way you design your industry, or whatever. You have to evangelise people; even when I came here 15 years ago, saying to some insurers “We can help you with some algorithms to better understand your risks”, I can tell you it was not easy. So, being the first on the market is not easy. But then maybe being the second one, just the second one, or the third one, but being the best, but having an idea that we match you enough with all the people and all the partners and so on, that’s where you’re going to be the winner.
05 Dec 2019
The imbalance of data between insured parties and insurers makes accurately assessing the insurance premium required to underwrite cyber risks...