Episode 41
with Thomas Krapf & René Alexander Papesch of Riskwolf

Nick:Hi, everyone, welcome back to another episode of Reinventing Finance, and together with my lovely co-host Tom. Tom, how are you this fine morning?

 

Tom: Thank you, Nick. Very fine.I'm looking forward to our talk.

 

Nick: And as usual, hopefully to no one's surprise, Tom and I have invited a guest, actually today two guests, Thomas and René from Riskwolf. Good morning, gents. How are you?

 

Thomas: Hello. Hello. Hi. Nice meeting you. Nice meeting and seeing you again, Nick. Nice meeting you, Tom.

 

René: Good morning, everybody. Just looking forward to the discussion now.

 

Nick: So without further ado, maybe why don't you guys tell us who you are and what you guys do?

 

Thomas: Yeah, maybe I get started. Just a short introduction. Thanks for having us, Nick and Tom. My name is Thomas. I'm the CEO and co-founder of Riskwolf. My background is from insurance, reinsurance and basically more than 20 years and I've been involved in a lot of like digital initiatives back already back in 2000, 2001. And this journey basically has been through more kind of commercial insurance, reinsurance, large scale insurance. And yeah, on the way to that, I've seen really different types of topics and challenges in the industry, how they utilize technology, data and digital means and when I met a couple of years back, René, my co-founder, and there we've seen the opportunities of what's coming up with cloud technology, what's coming up in the risk base, also how parametric works in the market already and what can be improved. And I think that is a little bit driving force why we built Riskwolf as a parametric service provider and technology. And yeah, happy to a little bit delve into this topic today, where we're targeting, what's the key trends and a little bit also, how we position Riskwolf in that.

 

René: Hi, my name is René, I'm co-founder and CTO of Riskwolf. And what I've been doing in the last decade or so was developing and architecting data intensive applications. I would also not want to phrase like big data and the other buzzwords. So basically when I left university, I wanted to look for a job as a data scientist, but back then nobody was looking for one. I think this has not dramatically changed and I have a consulting background and also a couple of years back, I met Thomas at one of this large digital transformation project in Zurich for one of the re-insurers. And basically back then we got the first touch points on the parametric insurance model that was quite fresh. And I think quite a new view on insurance that really got us really excited. And secondly, what we also have seen is like this kind of trend of that everybody's now new data becomes a way they were like a company starting satellites, capturing a lot of new data sources, but still all this new information is not fully leveraged in the insurance vertical. And with Riskwolf basically, we're trying to bridge this kind of a gap to really turn this kind of third party information into a workable insurance product using a parametric model now.

 

Nick: So maybe just to kind of dig into that before we kind of go into a little bit more broader, assume, you know, Tom and I work for an insurance company and we'd like to launch a parametric product, wrap our, you know, can you guide us through how Riskwolf is helping us along that journey? If that is a good way for you to kind of explain what you do as a business, if you feel a different way is better feel free, but so that we can get a grasp of what you do as a business as well.

 

Thomas: Yeah, maybe I can do a first, yes, get on that. So basically, as you know, it's insurance companies are also naturally mostly siloed by the different functions, different people involved. But typically what we are looking at is on the business side, is there a demand opportunity? Is there a client market they want to, they see as a fit for also like more automated solution and parametric solutions have two drivers. It's either the liquidity topic. So it's basically about, do my customer need immediate cash flows? That's more in emerging markets. Is that really a criteria which differentiates me in the market or is it something where I can really bringing in further coverage options? And the other one is more about looking at the mature markets. There are protection gaps due to the fact that in certain line of businesses, current capacity gets scarce or more expensive. So then all of a sudden parametric or is even excluded like in cyber. So all of a sudden a parametric solution becomes an additional rider or additional line item as well as a growth opportunity. So generally that's a key thing on having identifying the business market needs.

 

Nick: Just one question, because you and I think it's a nice way, you know, coverage gaps or liquidity and then you said liquidity is predominantly in developing markets. I was just curious as to whether you found certain appetite in large natural catastrophes, you know, big earthquake usually happens, lots of people and that usually puts, you know, payouts happen reasonably slow, et cetera. Whether you've also found at least that was the experience with kind of the recent, we don't, you know, have that many earthquakes in Germany, but we have some floods and, you know, it just takes a lot of time to kind of adjudicate these claims appropriately. So have you seen, sorry, it was a long winded way of asking the question. Have you seen that liquidity gap in terms of an interest in more mature markets as well? Or is this just me wishful thinking?

 

Thomas: Yes, maybe as again, from my end, Just in and also René can comment on that. We actually have positioned risk of a little bit in these kind of mid markets, not in the big programs, more like in the kind of embedded programs, like how can we support farmers? And there we look more in extreme weather scenarios and less in cat scenario, since there's a cat offering in the market, like with earthquake and the bigger insurers already have this. And what we see in the, a little bit on our market position, we go more into this kind of how now can BPP parametric brought a little bit to the smaller entities, to like states in India or to like programs where we can support those extreme weather events, like missing rainfall, too much rainfall, heating days, cooling days, and same thing happening also more on the income side. They're basically also a flood leads at the end of them to an income loss for a gig worker for a food delivery driver. So what we see is not about just a pure flood protection, it's more about how is an extreme weather event affecting people to make money or to gain income, or to protect farmers or people who have an exposure, like a property or farming exposure. This is what we see more and what we are focusing more with risk-proof. But that doesn't mean that the general cat exposure is going away. It's just already established more in the market. Whereas what we are going is more in this kind of additional weather impacts. That's an example. And naturally, also René, we can comment on those on other topics, because cat exposure is a known topic. And in my opinion, already quite good covered by the reinsurance industry.

 

Nick: And I, sorry, just to, for anyone listening, and because I didn't explain it really well, what I meant was an additional rider of immediate cash out. So, you know, you, the flood hits, you have your normal insurance that takes a long time, but maybe as an additional product, you get an immediate 20k. Right, which is based on some parametric triggers, because you have, you have, you need the cash for going to the hotel or, you know, getting someone to look at your house before the insurance kind of trickles through. That was kind of the idea that when you have these really accumulated large losses that you have an additional rider. But anyway, it's a minor point. It was just always something that I felt there is a liquidity, even in cat exposure, because it takes a long time to indemnify people. Okay. And so how do you help in that regard? What's your kind of, what do you do? What don't you do?

 

René: So basically when we're looking at new offerings or like new propositions to really get into new, it's basically three steps are involved. Like the first one, and this, I think the big difference to really explain us a bit, the concept is when you have a normal, like indemnity-based insurance product, it's quite important to understand the individual client profile or like the risk profile of individual customers, you know, to know basically how they behave and there's like a lot of, I would say where risk play a role also in terms of pricing and also understanding the risk. Whereas the big difference in parametric insurance is that we only look at like external events that are impacting customers. And the look earthquake is a good example, but it can be many more. And in a nutshell, if you have like robust information that you can really measure, then you can also really ensure it. And this is really about the trust. It boils down to this data source, basically, can we capture, for instance, earthquake information on the right level at the right granularity so it reflects basically on the client. So this is basically the main difference when you also trust from a thinking point of view, how you design this product. You need to understand basically what impacts the client from an external shock perspective in that sense. And the first step is usually, and this is also what we build up at risk for is exactly to model through that. So meaning it means when we take a good example is for instance, if you have like a business is operating, for instance, rental shops are, so rental shops, usually they would have a business interruption if there's heavy rain and they have no customers. So each day you can quantify them, nobody rents out bikes, whatever. And if you want to design for the coverage for that one, you would need to understand basically the certain rain patterns in a certain location where this business is located at. So meaning also here, the first step is really to get this in, defining the parameters, meaning what is the threshold of rain that really impacts this kind of nobody rents out. Is it a little bit, is it a bit more, is it a lot? And this is the basically driving the first step that's saying, okay, this is the coverage that we can then price and run through. So meaning you develop a price point and then typically carries gets involved basically where the carer can look at this risk and say, okay, I have risk appetite for rain for risk somewhere where this kind of rental companies are located at. So they can deploy capacity.

 

And the second step is what we then supported the risk of platform. We come from two fronts, basically on the one end is this kind of pre-priced coverages, for instance, a rainfall cover for this rental shops. And on the other side, you have the customer profile, meaning you have small shops, you have big shops and so on and so forth. Meaning the second step into the quotation distribution is then to calibrate this kind of covers towards the client size in that sense. And here we're talking about, for instance, what is the typical business interruption that they would, for instance, have for a given day? What is the maximum payout? What is the total sum insured? So these parameters usually like progress and ages can put in. And with this one, you can generate a quote. They can discuss it with the clients. If the rental shops like it, basically they would sign a contract, pay in the premium and then the interesting piece happens because in parametric, the conditions are predefined. That's a big differentiate  like there's a hundred percent certainty in that sense. So it's not about maybe, and you don't have this kind of gray areas. It's really about if the rain hits or in this contract, the rental shop gets the payment. And that's basically the last part of the platform that we take over is the kind of observation monitoring of all these different parameters. And you can just think of, if you have like a portfolio of companies, these are like millions of data points that need to be observed on a kind of regular basis. And once something hits, we would then generate a kind of a claim advice saying, look, these companies are impacted. There was an extreme rainfall pattern and we would send this information to the insurer so they can basically initiate the claim process. But that's the big difference. The claim process is then started not from the customer. It's really started from what's a neutral third party who looks at the data, who basically intersects this with the policy information, then calculates the actual payout that should be released towards the client. That's in a nutshell, the three steps basically that you need to do to have, I would say, a fully working parametric product up and running from an ideation state until to a kind of an operation state. That's basically what we support on the platform.

 

Nick: And out of interest, this sounds, and sensibly so, somewhat fit for purpose. It's not one size fits all. So at which point does a program make sense? How much premium are we talking to make it worthwhile, that effort? Are there some kind of minimum hurdle rates that you have encountered so far?

 

Thomas: Yeah, there's two key points here. One is naturally just maybe to reflect again what René said and what the parametric model brings. It's naturally in the ideal world, everything would be indemnity-based because then you reduce the so-called basis risk at the most efficient way at the most optimal way. But the problem that costs to do this last adjustment of policies. And generally, the smaller the policies are, the more automation you need in the claims process. Otherwise, it's not worthwhile to give out small policies. So that's the one driver. So the economic driver is at the end of the day, the cost of the claims process versus a little bit like the basis risk topic in parametric versus indemnity-based.

 

Nick: Can you just elaborate? I mean,

 

Thomas: That's one of the driving force, why there is parametric or why not?

 

Nick: Could you define basic risk? Because it's a very specific parametric term.

 

Thomas: Basically, if what René said, the measures need to be reflected to the loss situation. So the amount of rain, which basically brings in a problem with a farmer, with a yield of a farmer on a crop is basically something which has a relationship. But naturally, the basic risk is that you don't meet this relationship, that you basically pay out something where there's no loss or you don't pay out but there's a loss. So ideally, there's a correlation of the loss and the parametric measure, which is fully correlated. And that's what we cope with when you do parametric, what René said, when you do this calibration, we just find these tweaking points, which is, this is the amount of rain for this apple. If it's below that amount during the growth season, there's no apple growing because you cannot produce the rain otherwise. And this would then mean that you can check, then basically reduce the basic risk because the fact that there's no rain we can measure. And that means it correlates also to the fact that the apple is not growing there. And same happens if it's too much rain, then it leads to other problems. And same with wind, for instance, we can do and other periods where we have a clear relationship between the magnitude of a certain event and the problem it causes.

 

René: And the interesting aspect on the basis risk topic, typically a lot of discussions happens there in the parametric space. And it's normally really more when you talk about physical losses, because then it's usually, if you have an earthquake, it could be like, it's shaking, but maybe the building is not destroyed, then you have a basis risk or this would be more, okay, you give a payout to somebody who has and the building is still there, but the claim would be paid. But it could be also the other way around, or that, for instance, you define an earthquake somewhere quite far away, but you have basically an impact on your business interruption, or because whatever the city was closed, and you could not sell anymore. And there are like topics in both directions. And this is definitely true, especially for physical damage. But the interesting pieces of parametric is really to also look at beyond that, meaning if you really look into this kind of non-physical damage area, especially around business interruptions, or anything that is more cash flow related, where a lot of money at the moment is lost and not yet insured. And like an example here is when it's especially a technology risks, like for instance, interruption of or breakdown of internet infrastructure or cloud services, where you can just think of you have an online company operating. And basically, if your channel to the customer is down, you're not selling anything. So basically, you have quite a clear situation on the basis risk. And therefore, we believe, especially in this kind of more, let's say more dynamic and uprising new areas, there's a lot of demand coming in, the parametric model could be quite a good fit, because you have a lot of data available, you can really model that, you can define it in a kind of a parametric fashion, you can also bring it to the market. And that's basically also then, where at the moment, when you look at the insurance companies, there are not a lot of activities going on in this kind of a space. And then we believe that also with the parametric model, you can really get into this kind of new market.

 

Nick: And understood, I guess, one question, I think it was a nice way where you said, you know, you don't really need to understand the customer individually, or you know, as its kind of segment, we're kind of just looking at the perils, and then we try to reduce basis risk, but the perils need to be external. Have you found, and maybe that can segue us into the, because I think we've kind of implicitly covered pros and cons already, you know, there is a basis risk, but you have much lower cost of administration and

 

Thomas: Maybe one last one, maybe just because I wanted to make this point, you need also to compare with the existing insurance. So the existing insurance, especially in some farm crop insurance is naturally always difficult in terms of that. Typically, there's not a lot of payouts, the claim process is very lengthy, and a lot of claims are not adjusted because the form, the formula is not complete, or there's deadlines missed and then you, we should, one should also compare that the existing landscape of a lot of like indemnity based insurance programs is definitely not optimal for insurance. And I mean, And they look at it as a kind of, yeah, we get this insurance, but we don't expect anything out of that. And I think there's a trust topic, which, especially in emerging economies, we see that like in India and other emerging economies, there's a lot of missing trust in insurance. And what we see is the interest there is driven by people say, look, with a paramedic program, yes, it's quiet. It's more obvious, it is an immediate payout. And the payout is also triggered based on that information. And that's better than the current situation. And, and even taking into account some basis risk topics. Because what we do on the second topic, and then I shut up on that. Mostly, we're not, we're not selling direct insurance to farmers, to workers, we are going typically via intermediaries, that is people who aggregate demand, which are either producers of some crop, which have like 50,000 farmers or brokers or act tech companies who have basically, or gig tech companies who have already gig workers, SMEs, farmers in their portfolio, and who look at it as a product for their portfolio. So that is also your question, Nikolaus, you had about the sizing. There's no capacity provider, which gives an individual farmer policy. This is it has to be aggregated in the business case, which makes sense so if the premium amount, the target premium amount is below, I would say, 5 million US dollar, then typically no and since the parametric capacity is coming from typically large re-insurers, or really global companies, they have an internal benchmark. And if it's too small, they are not moving, because then it's too small of a margin for them. So it's definitely not interesting. I think there's a lot of opportunities to become more smaller and more granular to open up new markets for that but at the moment, I would say, there's a minimal amount of premium, which you need to showcase in the business case with your client. And that's also the operating models of risk. If you're not trying to get individual clients, we're trying to get aggregators, basically, who already have the distribution, control the distribution, own the distribution, who then bring in parametric as an additional offering to their clients.

 

Nick: Before we kind of go into some more use case, because I think it's fascinating for people to just kind of take some note and it's also important on the ideation or educational element of this. I was just curious before we kind of segue into that, because you've said external events. Have you found cases that are insurable around things like uptime as a service, ensuring SLAs of big producing companies that normally just do it out of cash flow, but potentially want to increase their SLA and kind of pass this on but it's not really an external event. It's kind of SLAs or their machinery not breaking down, one or the other one, things about increasing customer demand, where you price a certain flat rate, and all of a sudden, you kind of go, we think that's probably right, but we don't want to provide this flat rate in there. Or is this too close to moral hazard for your liking? Because no one's ordering a flood. I mean, we think we can all agree on that. Just out of curiosity, because you were so, and I thought it was a good definition of it's something external happening to you and this is kind of, I don't know, it's not black and white.

 

 

 

René: I mean, that's a good point. That is one of the key criteria is that you need to make a black and white decision, because otherwise we cannot have something that may be triggered. So meaning-

 

Nick: No, you can't. It's just not, it is black and white. It's just whether, how, would you say that a customer changing their usage behavior on some form of product, is that an external event that you don't have an impact? And if you have enough data about historic usage, you could create a parametric product and the same kind of for the SLAs on your own machinery and products, et cetera. Just out of curiosity, because the trigger is, I think you can set black and white, but it's kind of akin to, you have probably a little bit more influence on how you structure these things. And maybe there's a moral hazard about transferring these risks. I was just curious.

 

Thomas: Yeah. That's definitely one opinion is naturally, if you're looking more in the captive insurance market, actually you can build up monitoring with parametric means when also then the company with its own captive is also the risk taker and also uses its own balance sheet. And because they have no interest in to crash their own balance sheet with stuff internally, that would not make sense. So generally it's, again, still, I think as René said, we always start with objectively monitorable information. And this is not because we at RISK-Hooff are crazy about that, because the carriers want to have this objectivity because that's about the moral hazard. And I would say there's certain areas where it's definitely like in cyber attacks. If you want to ensure cyber attacks with internal information, then there's an incentive to launch your own cyber attack and collect the insurance premium, generally spoken. Now, naturally you would not do it this way, but I think there's certain areas which makes it a little bit more difficult because how do you prove then that is really external? So I would say as still as a point, if naturally the more standardized certain IOT measures are, and you can also observe it from outside, naturally you could say, I have my own IOT measures in my company and I utilize them also with an insurance policy, which is parametric, which observes that, but ideally it should be observable also from outside. Like that, yes, this has really happened. Yes. I would say that's definitely, I would say a gray area where some reinsurance do this kind of IOT offerings, naturally installing something in your house, like telematics in the car is also a little bit like an installing something in the car and utilizing this information from the car, which in general, you should not be able to, to manipulate as a driver. Yeah. That is, that is always the case that it cannot be manipulated. So there's a gray area. I would, I would say Nikolaus here, I would, I would not make a, but I would still be with  René, it should be binary. If it's something at the end, you have to then do the claims adjustment. If it's complex and you need an expert to find it, what really has happened, then you go into indemnity-based insurance because then you'll do a loss adjustment process.

 

René: Maybe just to add into this more on the requirements or what, what, what you need to have to really have a success case on that one. I think the, I think roughly three key criteria, one is what Thomas already mentioned is really like the credibility of the data source, meaning is the information really captured over time in a way that is not going to change. And secondly, or this one is, can somebody push a button and basically get a claim or meaning this kind of a trust manipulation. So meaning this kind of third-party data capturing should not be manipulated. So nobody should be able to trigger it or that's, I think one of the most important criteria. That's one really on the definition side. And then if you couldn't find a source and would fit in, I think then you already have, I would say already the first step done, but, and this is not in the discussion or happening with the carriers because secondly, somehow you need to come up with an insurance price or something. What does it cost? Like an SLA breach or what does it cost? Like a shortfall of visitors or whatever. And there are definitely a certain, I would say experiences are required still. So meaning still carriers are still past looking. So meaning to really like to want to see some, a couple of years data at least, or so meaning you should, then there's especially challenging for new data sources that don't have any history. So meaning there needs to be a certain experience already on, on the data. So they can look at that, what happens in the past and not really start back-testing that. That is the second criteria. And the last criteria is really more, I would say,  definitely something where the re-insurers come into play is anything around accumulation. Like earthquake is a good example or any of these bigger events, meaning that you typically have not only one customer impacted, normally, if one of these triggers are getting breached, definitely have like multiple people, for instance, in the same location being impacted, and that's basically the nature of this one, because if you have an excellent event, normally not only one is tied to that one, you have multiple people sitting there, meaning there's much more discussion about portfolio allocation, basically. Can you, for instance, on the distribution side, show up that you can really sell into the right, into the right split fence geographically or if you talk about technology risks, that not everybody that you have already risk in one of the data centers, or you want to split it across, but that's basically the first criteria to have like a success product here.

 

Nick: Understood. So, so what are the most exciting use cases that you think are currently in the, in the marketplace? You've, you've alluded to one, you know, we've done. Weather so so rain, wind, cyber, cloud downtime, we've said you know, any, what's, what, what makes the top 10 or top five list here?

 

Thomas: No, maybe, maybe just, I can comment on that. We, we also have started the risk group with more technology risk, internet downtime, cloud, but we've also seen that we have also pivoted 2022 into weather risk because we've seen the demand side is definitely on the weather side, because that's what the market understands, what brokers understand. There's already programs in the market. That's the typical here example is farming weather protection. There we see a lot of demand and that if I look at our current, more or less 80 active leads and deals, it's, it's, I would say of the, of the 75 percentage are weather related. This can be either on the farming side, affinity side, gig worker protection, as René mentioned, loss of income, but also we see now more captive topics like construction site, which cannot operate if there's heavy wind, event insurance where amphitheaters cannot open because of, and then you have a loss of income of, of, of, of the, of the performance, but also out of state travelers, which lose their ticket price and which lose the hotel. So that is, that is a little bit like the key drivers on the weather side. In addition, I think solar has come, come up lately, especially in Europe with all these solar installations. So solar on both ends and more on the yield end for the installation, for the bigger farms, but also solar, individual solar installations, how you protect your amortization of a solar installation on your roof against irradiation, like missing short, missing sunshine during summer, because that's part of your amortization naturally when you do that part partially. So these things, I think on the weather side is the major driver. I think maybe René, we can, you can comment a little bit on the cloud side and also on the alternative side, I would say that's two further categories, cloud technology, and then alternative is more like utilizing new types of measures, IOT measures, but also economic measures, like default rates of property as an economic measure for a rental company who has a lot of leases out and then identifies that as a, as a trigger. So there is a, there is, I would say 10% on our leads is on the cloud. 10% is roughly on, on the, on the alternative measures.

 

René: Yeah maybe just adding to the third category. So this is more, I would say that are not yet fully market ready, but I think this is really interesting also from the ideas, what is being discussed and what can be done, so maybe the kind of an outlook perspective and here I see definitely one, one area that is already started also on the insurance side is really around carbon, really carbon credits. And there, for instance, one concrete case that we got approached from a carbon trader is really about the, the, the project risks on the other end on like project delays, and then you have an afflustration, deflustration project and there are a lot of risks involved, like land cannot be rented. You have like a lot of the various input factors, but the consequences is basically that there are no forests, or that's basically the one that you could also put into kind of a parametric structure and, and start monitoring that with kind of remote sensing technologies and others. So I think the topic around carbon, carbon credits and really ensuring like more than the buyers of them or the project developers or the other ecosystems. I think there's a quite an interesting space in this one. I think that's definitely something where we also have discussions. And then there's another area, for instance, around like anything that happens in the internet or meaning not only down-times, but also like on the information side, like if you, for instance, take the example around the cyber attacks and then the consequence of reputation for companies or for individuals, there are, for instance, you have like a CEO account getting, taking over and causing a lot of reputation pain, meaning you see them like rumors are spread and just getting basically up and impacting like stock prices and this kind of topics where in the end you have the information. You know, okay. Somebody has a change communication profile. And then you also have information around, for instance, how does the company perform in parallel? And the same principle also applies of other segments. So they can just think about influencers or the selling online and all these other topics where it's more this kind of fake news and this kind of reputation and topics where at the moment, I think not a lot of insurance covers out, but if such an event happens, causes a big problem for corporates or for even for us, for individuals. So there's definitely a room I would say for parametric. And the third category, I think there are still something already being done. This is really around, I would say more as falling under the phrase, more nature-based assets. Meaning and how can you basically build up covers, for instance, around, I think that even currently from, I've seen some covers already up and running on like coral reefs and other assets that are really more false in this kind of nature-based area. We are also like a parametric model that can help project developers or like governments and so on basically to, to have rebuilding or getting this kind of assets up and up and running again. And that's, I think also something where you have a lot of topics in the kind of impact areas where parametric will play a role as well.

 

Tom: May I ask an additional question on the more on the business side of this. So did I understand correctly that let's say weather is more than 80% or something like that?

Thomas: I don't know. And if we look at our portfolio now and look at what is the sum insured and do you wait it to the markets we have again, we have Asia, Europe, and a little bit US, so still the majority request since we went live in Asia, we have more and more than roughly 70% of our business in Asia and actually there, weather is a big topic and there's brokers understand it, the market understands it. So it's something where the distribution already understands that is parametric can bring something to the table. That's, I think the driver, I think of that. If I look at the current portfolio, I think René mentioned it's more like a future portfolio. I would say the supply is missing. Yes. There's always like something which is innovative and there's always something where we already have.

 

Tom: How much is, how much is weather of your, because, because that's why I asked, did I understand it correctly? That let's say the biggest part is weather.

 

Thomas: Yes. 80%. Yeah. Okay.

 

Tom: Okay. Then I understood correctly.

 

Thomas: Yes.

 

Tom: I wanted to ask the business, more business type of question. What I, what you often see is that let's say people start companies and then you have clients and they ask, can you also do this? Or can you also do that? Um, but let's say if weather is 80% and everybody understands weather. Uh, let's say how much is your market share of weather? And just to ask in a very, perhaps a little bit blunt way, why don't you just concentrate only on weather? Because I think that's especially if you think for the future, that's probably because with all the climate issues, it's probably a huge market. Uh, why, why, why wouldn't you focus on that Pareto principle, so to say?

 

Thomas: Now it's a good question. Naturally we have, um, we have started, we have basically at risk of a little bit, uh, our perspective is that we will build up the weather because the market demands it, but we also will, I think from the way how it, how the future looks like in paramedic, that this type of technology risks will be the one where you need better monitoring. So what we have, our strategy is to support our clients on the weather side, but we will, we are never becoming the weather expert, like on flood risks and all the, so we are more like, we are, our vision is that we are becoming like more the, the, the, the, uh, the orchestrator of these type of solutions where then we bring in the right, um, the right, um, risk modelers, for instance, like for flood, because they're flood modeling modelers who are much stronger, who understand the flood and we cannot become that. I think on the, on the technology risk there, we have, first of all, we have, we have received a grant in Switzerland on, on, uh, from InnoSwiss on the Swiss innovation program, which is basically, um, in order to basically set up a cloud monitoring. And I think that is a little bit topic which drives us since that's where we see also the ability to enhance the market into new spaces and which is also where we think that's a strong asset of Risk-Hoof and, and, and personally, I believe at the end, uh, the world will be digital and will, and weather will be, will be an external component. So the intersection of critical infrastructure has also an intersection to weather events. So, so that's one and has an intersection to cloud services, all these non, non-physical losses happening in the digital world. So this intersection is an interesting one, which we believe we can, we can, we can really position  Risk-Hoof . Um, meanwhile, um, and we are a startup, we have limited resources. You're right. So we basically doing, doing, um, on the weather side, we're seeing that there, we can basically, uh, become break-even in the next 16 to 18 months with our activities, which at the moment in the current light of, of things in the VC market, it's, it's a good thing. So we trying to also build up this company, uh, in, in this way, um, that there is kind of a, a, a, a source of, of, of revenue and there is a topic where we think that is a source of future revenue.

 

René: Maybe just adding to this one as a more from a technology point of view, this was one of our key principles that we said, okay, we really want to build this kind of cross function. So meaning we want to basically have this kind of a middle layer that would sit on, on, on data sources. And this is somehow agnostic or does this basically the one you bring in like information, it can be rainfall. It can be an uptime information. It can be anything that you can really then feed your risk models and run through the process. So it's basically more just kind of horizontal approach, but definitely focus on like the first projects and so on. It's definitely on the market, on the weather area.

 

Thomas: And we can defend that much more. I think because insurance will change and the man you say, man, that's your podcast reinventing finance I think generally in insurance, everybody still does everything from end to end. And in, in other industries, it has started to be basically you become expert. Like Visa promotes has a Visa and MasterCard and other credit card. They do credit card transactions. And in insurance, everybody still does their own credit card transactions. More or less. I have a little bit over exaggerate this. And we believe naturally in future insurance also needs to be much more driven by components, market offerings that not everybody builds their own parametric engine. We built the parametric engine for the market. We go to market in weather, but we can accommodate other risks, which at the end of the day can serve as a utility or service for the market as a SAAS service. They can consume and they don't have to build their own. We also naturally are in a way also linked to this, to this destiny since insurance market. And this is why there's a lot of insure-techs at the moment, in my opinion, a little bit having problems. Insurance is a long-term sustainable business. You cannot cheat the market in being more clever than the market, even on the behavioral side and all these other sides. So, but what we can focus at RiskWolf, we can provide the market a strong utility or strong tool, which they, which, which also are in our opinion, will be much more used than in different scenarios. And, and then, and then we leave the market that somebody will come up maybe with the best flood model, or maybe three providers will come up with the best flood models. They should have the best flood models, not RiskWolf. But we want to be the legal component, which then somebody can put in these three best flood models, utilize RiskWolf as a, as a, as a, as a CPU, more or less as a, as a kind of a service, and then calibrate the products for their clients. Can be a group contract, an individual contract, a captive contract, a guarantee product, an embedded thing, whatever. A broker-driven market, an MGA-driven market, an insurance-driven market, whatever. We are, at the end, don't want to, we see that already in the market. We are active. We are active now in, in, in 20 countries with different leads and deal flows, and each market has an own way how it interacts with its clients. Sometimes it's broker-driven, sometimes MGA-driven, sometimes MGAs are not allowed. Sometimes there's, there's, there's a strong role of primary insurance. Sometimes there's a very weak role of primary insurance. So we don't want to become the ones who, who are the clever ones there. We want to be the, the tool set for these guys, for these markets.

 

Nick: If you, and I think you've already alluded to it, but would you have some predictions for the future of Parametric on how you expect the market to develop?

 

Thomas: Ah, yes. Data, and as René can comment on that, there's definitely on cyber side, you see that there's a lot of critical infrastructure exclusions that are coming since due to the capacity squeeze we expect in cyber, there's a huge demand increase and the supply on cyber, cyber coverage is probably stable or not really increasing the same way. So we see a lot of exclusions of certain critical infrastructure, which at the end of the day is the cloud downtown product. And there we see, we see a, we see a non, not yet, we see a demand side and economic loss side, which is substantially bigger than the current insured, insured losses, which is tiny, tiny. On the weather side, I think René can also comment here, and actually it's gonna be a topic which is not going away. And, and there we see the Parametric model grabbing just part of the weather risks where just you need to ensure a certain peril, a certain weather event. And, and really then being able to provide the market the right price signals for these weather events and then there, the Parametric model in our model has, has, has a future, not only in the cat, in the cat earthquake area, but also in this kind of extreme weather scenarios, which, which we believe with Parametric.

 

Nick: How do you differentiate just from a terminology, a cat event versus an extreme event?

 

Thomas: Yes, we, we, we initially said is when he started, yes, it's the same, but then we seen actually the carriers, they have different types of specialties on the, on the, on the, on the, on the, on the supply side, it's a different, it's a different type of, of risk. Extreme weather is different from a cat, a cat event is a, is a big flood, is an earthquake. An extreme weather event is a local deviation, like a certain rainfall during a season, which, which, which, or certain heating days during a, during a season. So the extreme weather is more like a local event, which, which has kind of rainfall, temperature deviations, whereas a cat event is really an observable one time big event, which hits like a earthquake.

 

René: That's like more low frequency and high severity, or there's a bit to differentiate of what, how the carriers look at that, but there's always a gray area in between or some, depending on the risk appetites or some, I think that's the definition I think it's a little bit fluent in some, some, yes. I mean, just adding to the, also to the discussion that Thomas brought up, definitely, I think the demand, also when you just look at cyber as a really, there's like really growing, everybody sees it. So definitely they'll be running much more into kind of a supply topics that there's less capacity or, so meaning really fulfilling this kind of, and then definitely like alternative ways will be maybe also seen much more often going forward, like ILS and others that also would fit quite nicely into this kind of parametric thinking. You can bring transparency also towards like investors along the chain. And then on the weather side, if you just look at whatever, what Swiss republishes on the new premiums being generated, like there's billions of billions or like in the next decade, I mean, this is a little bit, I would say they're more the long-term view, but also on a short-term perspective, what you also can do, you can also go, for instance, on the, like derivatives area on, for instance, on the, on CME, look at, at weather futures are, and they have, for instance, we see already that in like within one year, the open interest in the contracts that really triggered just from, if you just compare data 2022 to 2023, really see like a huge short-term demand. And you can say, okay, this is basically more a financial instrument, but the demand is basically the underlying pattern is the same, but just how it's delivered. This one is more through a hedge and the other one is for an insurance product. Yes.

 

Thomas: And CME, just to comment, CME is a Chicago mercantile exchange, which basically trades these kinds of commodity futures, which are already kind of standardized products.

But this is a sneaky signal that actually there is more demand towards weather topics. And the interesting thing is, and actually the trader products are only a subset of what, for instance, we could do at risk of, since they are limited to a list of cities in the U.S. And, and, and if you could then bring down actually this, this, these coverages, which are currently traded and add more coverages in, in, in, in, in, in smaller cities, in smaller regions, which are not covered, then that could be a pent up demand, which is, which, which could really be where paramedic could play a role, paramedic insurance, in addition to, to hedge products. Because it's, there's an intersection to that at the end, because every insurance at the end is a derivative. It's, it's, it's basically a put option where, where you, you get the premium and then you pay something which has no more value. You pay the car, more or less the value of the car if it's destroyed. So it's a put option more or less.

 

René: I think also like in 2024, we will also see much more that the parametric model moves out of the niche, just because like capacities just need to have alternative options, like really think also from just in the client conversations, they will just demand alternatives. Or if, if, if there's a, if there are parametric offering can be generated, then we will also see an uptake. We'll definitely also what, what, and this is, I think 2024 could be quite an interesting year for that. So yeah, across multiple topics, not only like, and our bet is basically we are going with the, with the, with the parametric trend or with our more infrastructure approach.

 

Nick: Gentlemen, before we wrap up, is there, is there anything that we haven't covered in, in this discussion that you feel would be a miss for anyone listening and, you know, wanting to get a, get a, get an update on what's going on in parametric? No, is a perfectly acceptable answer.

 

Thomas: Oh yeah, no, I think we had a, that's, that's not the different, different things you've covered. I think from my end, it's, it's, it's always, it's always I think innovative topics mentioned and there's topics in the markets and there's innovation with new topics. So it's, it's, I would say that needs to be differentiated a little bit where you can really bring it down. And, and I think in some areas there's still an education gap, especially on the broker side where we see we've had some meetings where the client was brought by a broker, but the broker could not explain the concept in the first two sessions and then we took over more or less from risk group and explained it and finally the client was convinced that's the right thing, but you had to go through one or two iterations. So we see that definitely that on the selling brokering side, there is still a gap in education, how, how parametric could help us a broker portfolio and not just take away their premium in the indemnity based portfolio, which they have already.

 

So that is definitely something where you see incentives on the distribution side to bring in parametric definitely something where which is also critical cornerstone to make this happening.

 

René: And I think just adding to this one is also what we also got asked quite often is why it's not yet fully, I would say common sense and people know it. And I think that's one is definitely the education or is basically the concept is not yet taught to the market or what Thomas mentioned, but still, what we see is that in the end, the whole knowledge of how you build this up, how you price it, how you underwrite it, it's different to the classic model. Meaning you need to have a different mindset of the people in insurance companies, because it's not anymore you need to go to a client and talk with them and see basically if there, if there's any dangerous situation, then can happen in a factory. It's much more, you need to have like data scientists really understands the events, can quantify them, can run them through. So that's basically you need a different skill set and also a little bit of different thinking. And at the moment, it's that just a handful of people who are usually sitting at the reinsurance and who only have very limited time to look at cases. And they only want to do the big fish. That's basically at the moment how it's currently done. And what we're just trying to do is a little bit with our approach to bring, bring this into technology. And then also have you already like first, first ideas up and running quite soon. So that's just need to be proved again on, on the carrier and saying, yeah, look at it. I've done this kind of checks. I'm confident with this. It's like 90% of my efforts. So basically please go ahead.That's basically also how we believe this could be really helped the supply side.

 

Nick: Cool. Gentlemen, thank you very much. Tom, any, any last words?

 

Tom: It's not perhaps it would be interesting to talk to each other. I don't know, one or two years and see how things developed because especially the climate topic will be much more in this whole common space with a very hot extreme summer. Then yes, it has changed in one or two years in a dramatic way. So it could be interesting to come back to you in the future.

 

Thomas: Definitely.

 

Nick: Awsome

 

Thomas: Thank you very much.

 

Nick: Thank you very much. Have a lovely day, everyone. Thank you everyone for listening.

 

Take care. Bye.