hosted by Inland Marine Underwriters Association. There are two ways to hear today's webinar. If you have computer speakers, you may listen to the webinar through your computer internet connection to the webinar in listen-only mode. If you lose your internet connection, you can still listen to the audio portion of today's webinar and follow along with the PDF of slides by dialing the phone number that was emailed to you with the login instructions. Today's presenter will have his video engaged, and it will be activated in a few moments. While your phones will be muted during the presentation, our presenter will address questions at the end of the presentation. Directions will be provided at that time. To view the presentation full screen, click the full screen icon in the top right-hand corner of the screen. To exit out of full screen mode, just click the full screen icon one more time. It's important to note that you'll need to toggle out to out of full screen mode to participate in polls and the Q&A. Today's presenter has several polls. To participate, simply click the appropriate radio button, or in a short answer poll, type your response into the text box and click the send icon right next to the text box, or click enter on your keyboard. Okay, we're ready to begin the webinar. There will be a short pause while we engage the recording. Hello and welcome to today's webinar titled, Cat Risk and Inland Marine, Earthquakes, Hurricanes, and Floods. I'm Lillian Coulson, and I'm pleased to introduce today's webinar. Please be aware of the IMUA statement. It's my pleasure to introduce today's presenter, Sam Lucas. Sam is based out of London and focuses on specialty lines within the model product management team. He is responsible for the marine cargo and specie risk, builder's risk, industrial facilities, and offshore platform models. Prior to working within the model product management team, Sam worked in the model specialist team, where he assisted clients with their fundamental understanding of the RMS model suite. Before joining RMS in 2015, Sam studied environmental geoscience at the University of Bristol, where he received a master in science degree. His primary focus was on atmospheric and oceanographic processes. Outside of work, Sam is usually found rock climbing, traveling, or trying out the latest food establishment. Welcome, Sam. The floor is all yours. Great. Thank you very much. Thank you for having me. Right. We've got quite a packed agenda today. Just to leave some time for questions at the end, I'm going to get cracking. The first thing we're going to talk about is what is cat modeling, just to make sure everyone's up to speed on that. Then we're going to talk about a bit more of the cargo and specie-specific items. By specie, what I mean is fine art, jewelry, or cat in transit. We'll be talking about the different considerations for each product and different storages, as well as the different peril considerations. At the end, we'll briefly touch on builder's risk and talk around that. Okay. Perfect. What is cat modeling? Cat modeling is the process of using computer models to accurately assess a given loss to a catastrophic event. The reason why this is important is shown by this slide. On this slide, we've got the top 10 insured losses from 2017. As we can see, the top 10 events on here, nine of them are natural catastrophes. One of them is the man-made catastrophe in September 2011, or 2011, I'm sorry. Of particular importance is, for the marine industry, Hurricane Sandy, which hit in 2012. As I'm sure you're all aware, Sandy is one of the largest marine losses of all time, potentially the largest. It really depends on what loss estimate you're looking at for the challenging explosion. With losses of recorders of about $3 billion, and it causes extensive damage to lots of different types of risk, from auto to fine art to different containers and yachts. It's highlighted the fact that understanding losses to different catastrophes is really important. It's really important to understand where your exposure is accumulated. Let's take a step back. The reason why we need catastrophe modeling is just down to the fact that they're quite large losses and are relatively infrequent. If we think about auto, in general, there are auto accidents most years. In 2014, there were 6.1 million reported accidents to auto. When we have that kind of level of detail, we can get an accurate average yearly loss to auto. If we start thinking about natural catastrophes, there's been a lot less natural catastrophes in that time period. If we think about from 1970 to 2015, there's been 4,905 different natural catastrophes. That's not just one peril like hurricane. That's all perils considered. You can't really think about it in the same way. You can't get an accurate average loss. To do that, we use computer models to simulate thousands and thousands of events that could happen to get an accurate view of what an average loss within a year would be. At this point, I am just going to pull out a poll. We're going to quickly ask you, what catastrophe models do you know of that are currently modeled? You should be able to see it on your screens now. Great. We're getting quite a lot of answers in. Well, you guys all got it. Yeah, in general, hurricane, earthquake, and flood, all modeled, including wildfire as well. So, I mean, obviously, last year was pretty bad for wildfires, so that's kind of highlighted the importance of that. And also, we've got one here, well, a few here showing terrorism. That's correct. Man-made perils such as terrorism and cyber are also modeled by catastrophe models. So, you're all aware of this. So, this kind of is a summary of what kind of perils are out there that we're thinking about here. So, before we move on and just talk about the marines specifically, we're just going to talk about what the catastrophe modeling framework is. So, first of all, the most important thing is we need to define our peril. So, what I mean by that is what peril are we looking at? Are we looking at hurricane? Are we looking at earthquake? Are we looking at wildfire? Once we've got that, we need to build a catalog of events that we could consider. So, if we're thinking about hurricane, we need to think about all the different paths that hurricanes could take, what the magnitude of those hurricanes are, whether it's a Cat 4 or a Cat 5. And then, because we're thinking about it probabilistically, we're not thinking about an event that has happened, which would be what we refer to as deterministic. What we're thinking about is stochastic, so probabilistic. We need to have some sort of likelihood of that event happening within a year to find for that event. Once we've got that catalog of events, we need to actually think about what we actually have. The model won't know that, so we need to enter that ourselves, and we want to have a specific view specific to our exposure. So, we need to put some sort of information in, whether it's address information like your street information or whether it's something like what the risk actually is. What product is it? How is it stored? Once we've got that, we then need to think about the hazards at the actual site. So, it's all well and good knowing what the hurricane path is, but what is our specific damage from that wind at that site for our specific risk? So, the hazard module will calculate that. Then, once we have that, we actually need to transfer it to an actual kind of damage that we care about. It's all well and good knowing that the wind's 110 miles per hour, but what has that actually done to our house? Has it damaged it, or in our case, auto? How much has it been damaged by? So, the vulnerability module kind of calculates that. Then, once we've got that, we have to actually transfer it into financial loss because that's what we're interested in at the end of the day. We need to know what our average annual losses are or something like our return period losses. So, for instance, every 100 years, we expect the loss to exceed $100 million or so. That's the kind of information that we want out of it, and that's the final step in the process. So, we kind of use cases of this. I'm sure you're all aware of this because I'm sure some of you use this in your day-to-day, but it can be used for pricing considerations. So, once you know what your average annual loss is in the year, you can start pricing your policies. You can use it for accumulation management. So, if you understand that you have X million dollars within a port, that would be really useful information for if there's a hurricane approaching that port, then you'd actually know how much you might have exposed to that hurricane. But then, there's also things like capital management. I'm not going to go through all of these because I'm sure you're all aware of this. So, quickly, we're just going to have one more poll because I appreciate this is quite a lot of information that I've just flown at you in a quick bit of time. So, just to make a checkpoint, is everyone with me so far? Great. Clearly, talking well. Cool. Thank you very much. So, moving on, what does this actually mean for cargo and species risk? So, it's all well and good to find in a framework, but we actually need to start thinking about this specifically to the risks that we're interested in. So, everything is going to be the same in the sense of the hazards and the events and the address information and the financial component will all be the same, but the thing that we need to think about differently is our vulnerability module because modeling something like cargo or species risk will be very different to modeling a property like a house. So, with that in mind, I've got four different products here that are generally covered within cargo and species risk. We've got bananas that have been stored in a container that might be temperature controlled. We've got a pile of gravel. We've got some artwork here, which is Van Gogh's famous painting, but we also have some auto. So, what we're going to do is we're going to have a bit of an exercise and just step through and think about what different perils could do to these types of risks. So, say the hurricane went through the site that we're looking at and there was a power outage. If we're thinking about this, the power outage isn't likely going to affect our fine art or our cars or our gravel, but it might have a considerable impact on our bananas in the storage because we might get spoilage. So, just knowing our product has kind of helped us understand, first of all, what our impact to our different products could be. Moving on, we're looking at the same site and this hurricane has actually caused storm surge, as we saw in Sandy, about three inches worth in this case. Then, it's a very different story. The gravel, there was actually some damage this time, but it's not a complete loss. We've just lost some of the gravel because it's been washed away a bit, but it's not the end or the complete disaster. For the bananas, if it gets into the container in the first place, it's one thing, but then it would only impact the bottom bananas that are within the water level, but it wouldn't affect any on the top level. So, again, we don't expect slight loss. With the painting, depending on the variety of factors, there may be some degree of damage, but the likelihood is the damage will be fairly limited. But for our auto that's been stored outside an open lot, we can expect significant damage. So, one final peril to consider is an earthquake. If we're thinking about earthquakes and we get the ground shaking, again, at the same site, for our cars, there's going to be minimal damage. Generally speaking, cars are designed to a certain extent with their suspension to be able to handle shaking, so we don't expect a huge amount of damage to these vehicles. For the gravel, again, minimal damage. For the container, they may fall over during the earthquake, and that could damage the bananas that are stored within them. That could cause bruising, and that could cause a loss there. And then we're thinking about fine art. Again, depending on the protection and the way it's stored, we could expect some degree of damage. Say, for instance, it's stored in the museum and the roof of that museum collapses, we could expect some damage to that art. So, the key thing here is just by understanding the product that we're looking at, it will make a vast difference to what our loss will be from the different perils. So, it's really key that we understand what our products are when we're thinking about each individual peril. Now, I kind of mentioned this earlier, but now thinking about storage itself. As you can expect, the way things are stored will make a big difference, again, on the loss for the same products. So, in this case, I'm going to go through the same kind of exercise again, but we're going to think about just also this time. And in this case, we've got also being stored in stack containers. We've got also stored in open lots, multi-story car parks, or stacked in kind of a car rack. So, first of all, what happens if we get that same storm surge again, that three inches? So, kind of similar story. Every other multi-story building, we wouldn't expect any impact because it's multi-level. It's not going to have anything stored in the ground in this particular instance. If there was some stored in the ground, then you'd expect some loss, but in this instance, we're saying that there isn't any car stored in the bottom level. Then on the car racks, the bottom level will be affected, but on top of that, it won't be. Same case with the container, unless one of those containers got moved and caused damage and the ones that are stacked above it. But for our instance with our open lot stored, also, there'll be considerable damage. Now, we think about this situation again, but for earthquakes, as discussed last time, also minimal impact in an open lot. But if we start thinking about these containers again, they might fall over, and then the vehicles inside might be damaged. We don't know how extensive that would be, depending on the amount of shaking and the amount of the containers moved. Then for the different structures of the car racks or the multi-story car park, depending on how that structure reacts to the earthquake, we might see much more extensive damage. For instance, we might have one of the floors of the car park collapse on top of the vehicles in one level, which would cause dramatic damage to that level, obviously. Then say if we have a tornado in this instance go past, again, to our multi-story car park, there may be some damage, depending on the level of damage to the structure itself. Same with the rack, but in this case, it's more open, so we'd expect higher damage than it would be in the multi-story car park. For open lot and containers, we expect quite significant damage because that tornado could pick those up and throw those around. So again, knowing the product's really important because once we know the products, we understand the different damage mechanisms that might happen to the peril. It's the same with storage. Once we know how things are stored, the different damage mechanisms will come into play again for the different perils. I just got some examples here. On a picture of one, we've got a temperature-controlled unit that's been damaged in some way. You can see the dent in here. In two, we've got the flooding and also we've got some containers that have fallen over due to the wind and the floodwaters themselves. In this instance, during the shaking within a warehouse, we can see that some of the racks have collapsed, which has caused extensive damage to the items that were stored on those racks. So these are kind of all things, different damage mechanisms that we're going to when considering the different storage types. And just to briefly mention, container vulnerability, there's not a huge amount of research that's been done about the kind of damage we can expect from containers being knocked around or falling over at a site. So one of the things that RMS has done is produce simulation-based analytics, which looks at how containers might react to an event that causes shaking or wind or waves. So just to demonstrate this, we've got some on the top. We've got our containers being knocked around by wind. We can see that the purple containers are the ones that have been moved by the wind. We also have some containers in the bottom that haven't been affected. But for surge, again, the whole bottom layer will be affected, and then there might be some more partially submerged containers as well. And the same sort of thing for if debris has been picked up by that surge. So for instance, in a lot of instances we saw in Sandy, when the surge wave hit, a lot of those containers were picked up or the auto themselves were picked up. Those things swept across and caused more damage to the other risk within that port, so it's really important we consider that as well. So at this point, just to quickly recap it all, we're going to do one more poll. This is the last one, so I'm not going to ask too many questions after this. You can ask questions after that. I just want to ask quickly, which products do you think would be more vulnerable to wind and surge risk in Miami between also electronics, petroleum, and fine art? I think that's quite a few most people now. I'm just going to broadcast the results for everyone so everyone can see what everyone picked. I mean, if everyone wants to vote, do now before I broadcast. Okay, perfect, broadcasting now. So what we can see is that we've got 57% of people who voted for auto, and then followed by electronics, fine art, and prosodium. So I mean, it's a bit of a trick question, really, so it's a bit unfair of me to do this because it does depend on the storage. I mean, if there's fine art that's just lying around in an open lot and there is some flood water, obviously, that would be catastrophic. But in general, what we can consider auto would be the most vulnerable to wind and surge risk in general. And then this is followed by electronics, then plutonium products, then fine art. But again, within that, depending on the storage configuration, we can get a lot of relativity in it. So for instance, electronics that are stored in containers within the warehouse would be less vulnerable in some instances than plutonium products. So this just kind of highlights the fact that it's really important. We do need to know the storage and products. Okay. Perfect. So now we've talked about that, we can start thinking about other features. So we kind of refer to these as secondary modifiers. So if we're thinking about protection measures that are in place, I'm sure you're aware hail causes havoc with auto. And so in a lot of instances, we need to have hail nets. So having a hail net will make a massive difference to not having a hail net. I wouldn't want to buy that car myself without a amount of hail damage. So it's kind of important, if we have information like hail nets, we can also capture this. And it's important we do think about that, because the losses can be substantially different. And it's not just hail nets, we can think about other protection measures as well. So for fine art, we could think about it being behind glass or something as well. And then the other thing as well is with cargo and species risk, there's a lot of salvage potential within it as well. Just because it's been affected by wind doesn't mean it's a total write-off. For instance, if a jewelry is knocked off some stand, assuming there's no damage to the jewelry itself, the salvage potential of that is 100%, so you can just pick it up and move it along again. But going back to the auto example, if we think about the salvage potential between a Ferrari and a general market car, like the Peugeot we've got here, the salvage potential of the Peugeot is going to be a lot greater than that of the Ferrari, partly because of the brand clause protections that might be in place to limit any kind of salvage potential, but also stuff like the replacement parts. We can expect that the replacement parts for Peugeot vehicles would be a lot easier to find than replacement parts for a Ferrari. If you do find replacement parts for a Ferrari, then let me know, because that would be great. And then the other thing we can think about as well is packaging. So obviously, it being cargo in containers, it's going to be packaged, it's going to be wrapped up. And that's important that we can consider that as well, but obviously the degree of packaging can vary substantially as well. So on this slide, I've got some kind of examples of how bikes are packaged and transported. So in one example, we've got a bike that's being stored in cardboard. Say that there's some wind, that cardboard, sorry, blood water, that cardboard would just melt away, and then the actual bike itself would be damaged. And then if we think about a more professional container, like we've got on the right, which is a bit waterproof, the damage would be very, very different. So as well as the product and the storage, these three other kind of features, so the salvage, the packaging, the protection, are all equally important and can make substantial to your view of risk. So we've kind of gone through the characteristics of the risk itself. Now we're going to think about the address information and why that's important. So by its nature, cargo is moved around the world, as I'm sure you're aware. I mean, in this example here, I've kind of done it on a global scale, but if we think about it in Japan, the highest risk in Japan would likely be earthquake. Say that this car wasn't exported to the US, then if it landed within San Francisco, then the earthquake risk there would be higher than, say, the hurricane risk. But then if it was transported to somewhere else, like, for instance, Florida, then the hurricane risk would be a lot greater than the earthquake risk. And then as we go inland, we can then expect tornadoes and hails to become more important risks than, say, earthquake or hurricane in that instance. So just understanding a bit about where your cargo is being stored in each given destination will make a massive difference on the peril that will be the driving risk at your site. But if we think about it on a more granular scale, specific information can also have a major impact. So what I've got here is a map of the port of New Jersey and New York, and what we're doing here is we've modeled Superstorm Sandy and looked at the kind of surge losses, wind and surge losses, throughout the terminals. And what we got displayed is kind of the percentage of loss that each terminal took up. So we can see here in this darky pink color that the Port Jersey Port Authority Marine Terminal makes up 55% of the overall loss. So say if we knew that our specific cargo was stored in this terminal rather than the Elizabeth Port Terminal, which took up less than 0.5% of the overall loss, then we can see quite substantial differences in our view of risk, which will have a massive impact on our pricing, our capital reserves, as well as our accumulation management in itself. So the more granular the information you have on the address front, the more accurate your view of risk will be, especially for high-gradient perils like flood. And what I mean by that is surge waters can vary dramatically across a street. Other things to think about as well is the fact that, by its nature, cargo isn't stationary. So far, the catastrophe modeling framework that we mentioned earlier is thinking about losses from an annual average. So we need to consider how long that, A, that the risk is dwelling at these different warehouses, and B, then consider how many shipments per year we might expect for that warehouse and terminal. So on this plot here, on the y-axis, I've got loss, and then on the x-axis, I've got the return period, so saying, say, this point here was we get this amount of loss every 50 years, or we can expect to. If we then think about the risk being stored for only a quarter of the year, then our loss obviously lowers, and then if we then think about only a month, then it lowers even more. So it's important that we then also understand how long, typically, these risks are staying at each site, and that's something that is important to capture, because, again, as you can see, it makes quite a big difference to your view of risk. The other thing as well is cargo, I'm sure you're aware, exposure will vary quite a bit for one site. So within a port itself, you can imagine that just before Christmas that the amount of exposure within ports would increase to meet the Christmas demand, but this will vary in different regions, and so having some idea of how the exposure will vary at key times of year, for instance, in the hurricane season, a lot of ports will move a lot of their exposure out during the hurricane season, so having that kind of information is quite important as well, because that will have a dramatic impact on the actual value within that site. So to collect everything together and think about things in a bit more of a complicated way, we can start to think about stock throughput policies. So the example I've got here is wine that's being produced in Mendoza in Argentina, which is the red dot on the map. Then this wine's being shipped from Mendoza to Santiago de Valparaíso, one of the ports in Chile. From there, it's being shipped across the world to Japan, but also into New York. Then from New York, it's then being taken to several different sites within the US, or maybe even across the border into Canada, and we can start to think about this in a bit more of a clever way. So say we know where the location is within each of these countries, we can then figure out what peril would be the prime factor, so outside of the US, it's more dominated by earthquake, but then in the US, we can think about earthquake and hurricane. We know how things are being stored, and we know what the product is at each of the stages. So for instance, within the vineyard itself, it might be consumables, because it's the grapes themselves, and then as things move on, the product itself changes, and we can think about that. And also, a vineyard will react very differently, because the storage of that will be very different to, say, a distribution center. So say we collect all that information, and we use different considerations, like it's a temperature-controlled good, it's consumable, it's being stored in liquid tanks, it's being stored in the warehouse, we know what the values are at each of the sites, and we know typically how long it stays at one destination, and how many shipments they get within a year, we can start thinking about the losses. And so here's an example for the different locations we have, and what we're comparing it to is the marine model to a view where we haven't considered the specific storage and product considerations, or the specific time considerations. And the key thing here is that using the different product information and storage information as well as the different time information, we've just got a more accurate view of risk. By location, that view might see an increase in loss, but we might also see decreases in loss, the key being that it's more accurate. And at the end of the day, we can then start thinking about these stock throughput policies in a slightly different way than we are currently. So to kind of summarize, what you're insuring is really important, because that will have very different damage mechanisms for the different products. How that's being stored is very important, and that will have a massive impact, again, on the different damage mechanisms. So whenever we think about our additional factors, like our storage, like our protection and our packaging, and then where it's being stored will have a big impact, not only on the more broad level, for instance, where our more, or what parallel is the biggest driver, but also on our more granular level, like within our terminal itself. And then when time it's stored at that place, and the likelihood that the product will be there at that given time is really important. So overall, this is kind of the framework that the RMRS marine model uses to assess cargo and specie risk from catastrophes with 18 different prototypes and 12 different storages. So that's your what it is and how it's stored. And then going to where and when using the different features, as well as additional things like protection and packaging. So I appreciate I whizzed through that, but that's kind of the cargo and specie section at the end. I'll take some questions on that at the end, so just if you have any questions now, just hold them for now. For now, we're going to switch our kind of thought caps a little bit and start thinking about our builder's risk, so our constructional risk or erectional risk policies. And we do need to think about these in a slightly different way, but the key is the framework is exactly the same. The only thing that's different is we're now thinking about how the hazard damages these locations in a bit more detail or a bit differently. So the key thing with builder's risk is that obviously the value changes with time. When you first start constructing these projects, you've got the foundations. Obviously, there's going to be a lot less value than at the end of the project, whether it is complete, but it's unoccupied. So when I'm saying phase one, I mean foundations. And then phase five is complete but unoccupied. And then phase two, three, and four is in between, where phase two is where we start the construction, the building envelope, but it's not finished. We need to consider that, the value ramp up itself. We also need to consider how that vulnerability to that specific part of the project will change through time. So if we're thinking about Hurricane, you might think that when it's fully finished, it might be at its most vulnerable. But in general, that's not the case. But in general, that's not the case. What we've found is generally in phase three, when the building envelope hasn't been complete, and what I mean by that is the walls are all up, the roof's on, but the windows might not be in place yet, and there's still openings within the building. So when that wind and that rain gets within that building, you can get extensive damage to the structure itself or damage to the roof because the wind can then go into the building and blow the roof off. So understanding the different vulnerabilities at each phase is important, and then combining that with the value ramp up, you can get quite an accurate view of risk from these construction projects. So just to show this a bit more graphically, on the y-axis, we have the project value with time, and on the x-axis, we've got the project time. So typically, we might expect that within a year, it takes about 20% of the time to finish laying the foundations, and then each phase will take a different amount of time to complete, depending on the complexity of the project. And just to kind of build out that definition of phase one and phase five in a bit more detail, because it's not the most clear, but we're thinking about a heavy industrial building. Phase one will be like your land clearing, your site preparations, and the installation of your foundation. Phase two is direction of the first bit of the structure. Phase three is we've installed some of the equipment, as well as the key structures. Phase four, we've then started thinking about the detailed work that got into the equipment, the detailed work that got into the equipment, and the different testing that's been done to the structure. And phase five is it's finished and is currently unoccupied. So once we combine all this together, we've got our very different views of risk. And this is an example to earthquake, where phase five is the worst instance. So going from the cargo and speech model, where we've been thinking about our different products and peril considerations to the different products and storage, we're thinking the same thing here. But in this case, we're thinking about how the occupancy will affect the different equipment within the building and the different roof types that might be on these buildings, but also the construction itself, whether it's wood or unreinforced masonry. And you can have quite dramatic differences. So to summarize, what we've seen over the years is catastrophe losses do need to be understood. What they are, the largest drivers of insured losses and the largest marine loss within Hurricane Sandy, it's important that we do understand catastrophe losses. To understand these risks, we really need to know what the risk is, whether it's an unreinforced masonry building that's for commercial use, or whether it's our auto being stored in an open lot. Then where it is will have a big impact on the different peril that we're considering, but also have a big impact on the hazard we might experience at that site. And the key thing is the more information you have, the better you're able to differentiate your view of risk and better assess your accumulation to that risk. Great. Thank you for your time. I'm going to open it to questions now. Okay. Thank you, Sam, for an excellent presentation. We're bringing the Q&A pod up now. At this time, our presenter will take questions. To ask a question, once we get the Q&A pod open, you'll type your question into the pod and click the balloon icon next to it. And we'll wait a few moments for questions to come in. We have our first question. You say that capturing data is important, but are there any specific bits of information we should be focusing on gathering in particular? Yeah, great question. Thank you for that. So I mean, in general, I understand it's probably the best case, but the more information you have, the better. The more information you have, you know, the better view of risk you're going to have and the better differentiation you'll get. If push comes to shove and you can't get all that information, the key kind of information you can kind of focus on is for marine and, sorry, cargo and species risk would be the product and the storage type as well as where exactly it's located, but on the builder's risk side, it would be more kind of what is it constructed out of, what type of building is it, and then where it is as well. So in all cases, location will be really important. Thank you. We have a... So one of the attendees said, thank you for presenting today, and can you go over the difference between the calculating damage and quantifying financial loss in the model framework? Yeah, sure. That's a great question, actually. Sorry, I appreciate it. It was quite a broad overview of the catastrophe modern framework there. So the way I like to think about it is in the hazard model, we've got... What we've got at the end of it is a wind speed for our location, if we're thinking about hurricane, but if we're thinking about a building, we'll have a kind of replacement value and we'll have some exposure value that's... We'll have a replacement value that's in there, and so what we'll have is called a mean damage ratio or an MDR, and so that will look... That will be a link between the hazard and a damage to its structure, and that could be through a kind of simulation or it could be through kind of looking at claims history and seeing what damages we've seen from a specific wind speed. And then once we get to the financial risk component, we take the actual exposure value itself, so your total insured value, so say it's 10 million, we apply that to the mean damage ratio, and we'll get our kind of financial loss, and then from there, we then can think about the different insurance terms and reinsurance terms that might apply upon that location. Does that make sense? Perfect. We have another question. Is Lloyd's using the same modeling approach to catastrophe? Yeah. In general, I can comment from the RMS perspective, but in general, the catastrophe modeling framework is fairly consistent throughout the market. We work directly with Lloyd's on their regulatory stuff and help them pick their realistic disaster scenarios or their RDS events. So yes, they are using this catastrophe modeling framework. Another question. What impact has climate change and global warming on cat modeling? That's the big question. So I mean, I can talk briefly on it, but I'm not going to be able to give you an exact answer because I'm not a climate scientist by trade. But what it will have an impact on is the frequency of events as well as the general hazard itself. So say, for instance, you have warmer sea surface temperatures for hurricanes, hurricanes derive their energy from the heat. So the hotter, the warmer, the higher the energy you have for a hurricane. So the faster that the speed of the hurricane can build up. So that might not impact the frequency of it, but it might impact the intensity of that hurricane. So you might expect there to be more cats four or five hurricanes. But again, that's just a very simplistic view of it. There's a lot of other things that fall in that. And we do... At RMS, we consider that using prediction models of the different climate states over the next five years, which we call our medium-term rates compared to our long-term rates, which are only looking at the historical average. Thank you. So we have some questions about other forms of inland marine, such as contractor's equipment. And how... So what is your approach to modeling, say, contractor's equipment? Do you have more specific models that would break down into those various lines? So it depends on the particular type of equipment you're thinking about. In general, the framework will be the same. So just as I talked about with builder's risk and marine, the events and the hazards will all be the same. The bit that will be different will be the vulnerability component itself. So if you're thinking about cranes, then we do have specific, what we call vulnerability curves that describe that kind of damage. But if there's other equipment as well, then it really depends. So I can't give you an exact answer, because it really depends on what equipment you're thinking about. But, I mean, if you are licensed and using RMS, then you can reach out to us and we can help you kind of have a look at that equipment and what might be the best approach to modeling that. All right. Thank you for an excellent presentation.

marine risks

data assessment

risk assessment

financial loss

perils

catastrophe modeling

inland marine risks

vulnerability curves