Sound Policy: The growing cost of hail

Brian Amaral  
Welcome to Sound Policy, a podcast from FM where we bring you the latest insights on what it takes to protect today's businesses. On today's show, we'll be discussing hail and a new map that shows where the risk is most severe around the world. To break it down, we're talking to Andreas Muehlbauer and Anthony Longabard, two experts here at FM who've worked on this project. Anthony, Andreas, thanks for being on the show. 

Anthony Longabard  
Thanks for having us. 

Andreas Muehlbauer 

Awesome. Thanks for having me.  

Brian Amaral 
Anthony, Introduce yourself. What do you do at FM? What's your role? What's your day to day? 

Anthony Longabard  
So, I'm a senior staff engineering specialist within the Chief Engineers’ Group in the Climate and Structural Resilience Team. Uh, one of the things I do is I, I manage a lot of our data sheets that are associated with structural engineering, and that's a, a wide berth of topics of which hail is one of them. 

Brian Amaral 
Data sheets for the uninitiated. Briefly, what is a data sheet? 

Anthony Longabard  


So the, the FM loss prevention data sheets are basically FM's home of where we provide all of our recommendations to our insured clients as well as the world at large, because they’re free
to the public. So, encompasses all of our years and years and years of experience, from a science standpoint and from a loss experience standpoint. So, we have an FM property loss prevention data sheet. That's 1-34. That's hail damage. 

Brian Amaral 
Are you the author of the hail one? 

Anthony Longabard  
I am now the caretaker of the… 

Brian Amaral 
The caretaker. 

Anthony Longabard  
…of the hail data sheet. It's been around for, in some form or another, for a very long time. You know, we've been insuring hail for, I don’t know, about a hundred years or so. So, it's been around for a while. So, to call myself an author is a little bit ambitious. 

Brian Amaral 
You're a, you’re a, you're a steward. A steward of the… 

Anthony Longabard  
That's an excellent, that's an excellent way to describe it. I'm the steward of, of the data sheet. 

Brian Amaral 
Andreas, how about yourself? Can you introduce yourself? What do you do at FM? 

Andreas Muehlbauer 
Yeah, sure. So I'm a principal research scientist here at FM in the Natural Hazards and Climate Risk research team. And my background is atmospheric science and climate science. I've been with FM about 11 years now coming up, plus or minus. And I've been the lead developer for the global hail map and the model that's underpinning it. 

Brian Amaral 
You've been involved in climate science for about 20 years now, I understand. How do you think the understanding of the hail risk has evolved in the 20 years you've been involved in climate research? 

Andreas Muehlbauer 
Yeah, that's a great question, because 20 years ago, hail wasn't on the radar for most insurance and reinsurance companies because it didn't generate the big losses that we see nowadays. And obviously, hail is one of the perils that's been trending upwards greatly. And we've seen this in natural hazard losses worldwide. 

Brian Amaral 
Why is that? 

Andreas Muehlbauer 
Well, it's really because of trends in the underlying weather systems that create hail in the first place. It's not entirely fully understood because there's varying factors contributing to it; but we've just seen in the losses, increasing frequency of these events. 

Brian Amaral 
You work on a lot of different things. Hail is just one of the risks. Where do you place that on all the things that, that from a climate perspective that are facing businesses right now, from an insurance perspective, whether that's tornadoes, straight line winds, hurricanes, flooding, where does hail sort of rate in that? 

Andreas Muehlbauer 
It is not the biggest factor, or maybe not the one that's like, um, controlling the headlines of a news article, right? Because it's not causing homes to collapse or be in fire, right? And people losing homes and things like this like we saw in LA last year. But it's more a constant, constantly increasing background peril that, you know, hits us more and more and more often, right. The cost and the losses have doubled in recent years, and the trend seems to be continuing. 61 billion was severe convective storm peril that includes hail. And actually hail is the, the majority of the losses in 2025. And we had a very similar amount in 2024. And what's even more important in this context is that historically, and historically means the last 20 years or so, industry losses have been somewhere between $20 and $30 billion. Now, you're talking about a factor of two here. 

Brian Amaral 
Yeah, doubling. The, the cost is doubling. 

Andreas Muehlbauer 
Exactly. 

Brian Amaral 
And, and Anthony, you've been at FM for some time now. How has your understanding of hail evolved? Is there something that's really surprised you as you've gotten more into this? 

Anthony Longabard  
Um, I would say I'm not surprised. You know, prior to my time within the Chief Engineers’ Group, I was with FM Approvals, where we did a lot of hail testing of different products. So, I was, I was aware of the significance of it from a testing standpoint. We, you witness testing, you realize that it's, it's a pretty significant hazard. 

Brian Amaral 
Describe what does it look like? What do those tests look like? 

Anthony Longabard  
It's a very dramatic test. The most interesting one is anytime you're testing a very severe hail. So, in all cases with very severe hail, we're using an ice ball that's shot from a cannon into a test chamber where whatever it is you're testing is sitting, and it gets struck by that, that piece of hail. It's quite something. And then to see the aftermath of, of something that fails. Roofing systems, skylights, photovoltaic panels. Those are the big three. And it's extremely difficult to survive a, a hit from something like that, particularly anything that happens to be brittle. 

Brian Amaral 
Andreas, let's take a step back. How does hail—from a climate science perspective—how does hail form? Like, walk us through what a hail storm actually is and how it forms. 

Andreas Muehlbauer 
So, first of all, you need the right conditions. You need typically warmer temperatures or a temperature gradient between cold air and warm air, right. Something that creates buoyancy. So, uplift in the atmosphere, right. And then you need water vapor. Water vapor is the main ingredient. Once it gets lifted up, it reaches the level that's called condensation level. It starts forming cloud particles to be starting with, with small liquid droplets, right? These liquid droplets collide with other droplets, form bigger drops. Those bigger drops get still lifted up to the freezing level. They start freezing. Those frozen particles start aggregating other frozen particles. They become bigger and bigger. Um. More lumps, right? And this process continues until the particle is so big and, and so heavy that it can't lift it anymore. And then it starts falling out of the cloud and eventually reaches the ground, if it doesn't melt. 

Brian Amaral 
What is it, like, look like on the ground? Paint a picture for us of what a hailstorm looks like as it's forming, as it's rolling over an area. 

Andreas Muehlbauer 
The classical ones are the very huge, we call it supercells, that form in, for example, Oklahoma and Texas, that are really tremendous, and they are also beautiful at the same time. If you look at them, um, at the bottom of the clouds, you know, often look like UFOs, you know. They have these rotating, very smooth layers of clouds, you know. And that rotation is also what, what generates the vigorous updraft that then also generates the very large hail. that they can cause for a moment, right? They're actually very beautiful if you if you slice them up. If you cut them with a saw, you can actually see there's different layers of ice in the stone itself that are characteristic of the physical processes that were involved in creating them in the first place. 
Some of these layers are very opaque and others are very clear. So you can kind of see the differences in the growth stages of the stone. And you can typically also see a little seed in the middle at the center that, that was the, the initial seed, where basically it started growing around. 
So it's very fascinating. And they have all sorts of weird shapes that are just a reflection of how these particles grew in the cloud. 

Brian Amaral 
Anthony, from a structural perspective, what sort of damage does hail cause for, for businesses around the world? 

Anthony Longabard 
So, from a structural standpoint, I mean; it's not a concern from a loading perspective. But from an impact perspective, it causes a lot of damage. So, you end up seeing a lot of, you know, roofs and other equipment that end up with dimples. It doesn't necessarily affect the, the service, serviceability of what it hits, you know, with the exception of things like roof mounted equipment. Um, but really it's the after-effects of what stuff looks like. You know, if I happen to have a metal roof and it's all dented, you know, it doesn't look so good. You know, for a for a client to have a roof that now appears damaged, even though it's still serviceable. You know, you really want your, your, uh, roofing system to appear as if it's undamaged, particularly if it's a high visibility sort of situation. You know, you might happen to have a roof that's part of a headquarters, or you have a lot of clients coming to visit your facility. Having a facility that is aesthetically damaged is not exactly pleasing. 

Brian Amaral 
What separates a roof that can survive a hailstorm versus one that might end up with holes in it, with causing leaks, those sorts of things? 

Anthony Longabard  
Um, you know, that's a, that's a tough question to answer because it really depends, system to system. Um, for roofing, one of the things that can affect it is, uh, it's the, the thickness of the, the membrane layer. Uh, it's substrate, and if you happen to have anything like, uh, insulation plates that happen to be between the two. You know, you have a piece of metal that's sandwiched between a fairly flexible layer and a rigid layer, the chances of that tearing through the, the, the roof cover layers is greater, so um. But again, I mean, over time, over the last ten years or so, there's been a lot of evolution in roof systems since the introduction of VSH. From, uh, FM’s standpoint, manufacturers have done a lot of work to try and find systems that can survive this kind of impact. 

Brian Amaral 
VS...VSH. What is that? 

Anthony Longabard  
Very. So, that's very severe hail. So, that's the high end of um, what we have for mapped hail intensities. 

Brian Amaral 
Let's, let’s, let’s get into that. The map. We, FM recently announced the launch of a worldwide hail map. Um, Anthony, starting with you, what was your involvement in the, in the new hail map? What, what did you work on? 

Anthony Longabard  
So, what I worked on, Andreas and his team basically developed all of the, the maps themselves. Um, my work on that was basically trying to take all of the excellent work that they did and, and translate it into our, our data sheet itself. 

Brian Amaral 
Andreas, what was your role in the, the hail maps? Walk us through what you did and, and,,and what these new maps are. 

Andreas Muehlbauer 
Yeah. So, I was the lead developer of the global hail model, and that ultimately feeds the global hail map that we have nowadays. And it was really a multi-year process, where we started out just with hail observations all around the world to kind of produce hazard maps, purely from the observations themselves. So, basically, looking at the frequency of occurrence of hail storms and basically also how, how large the hail is once it hit the ground, right. But ultimately, we couldn't reach a global hail map with this framework because hail observations don't exist everywhere. You have very good coverage and a good history in the United States. We have a still evolving history in Europe. We have some observations in Australia. But for a lot of the, the planet, you know, there's really not a good concerted effort for hail observations that are good enough in quality and long enough in history to really come up with the reliable estimates of the risk. So, what we did then is basically we combined our previous approach with, um, what we call storm of environmental parameters. So, these are, um, atmospheric parameters that have historically known to the scientific community to produce, um, large hailstorms, right. So we, so we combined that approach, looking at the environmental parameters with the previous approach of looking just at the, the hail observation itself. Combined that together, um, with a machine learning framework, and that allowed us then to actually also predict the hail risk in areas where we traditionally didn't have enough hail observations from the past. So there used to be only US and Australia hail maps. But now with the new approach combining historic observation with environmental parameters that are known for producing large hail, we can also produce a worldwide hailmap. 

Brian Amaral 
Machine learning. Could you explain what that means? 

Andreas Muehlbauer 
Yeah. Machine learning. That's, that’s basically just a fancy word of, of, of saying we developed a statistical algorithm that could connect the dots between atmospheric parameters that are basically responsible for this thunderstorm development and the hail that we observe on the ground. 

Brian Amaral 
What does this map have that other ones don't? 

Andreas Muehlbauer 
Our map really quantifies the risk in terms of frequency and intensity. So, how often hail occurs and how big the hail particles really are, right. The hailstones that hit your assets. Other maps don't do that. They only look at things like number of days where you have hail occurrence. Well, that's great to know, right? But it doesn't tell you anything about the loss potential of the hail that's falling. So that's a, that's a key differentiator. 

Brian Amaral 
Yeah, big, big difference between a marble and a grapefruit, right? 

Andreas Muehlbauer 
Correct. Yes. 

Brian Amaral 
In Europe in particular, what are some of the areas that that jumped out at you when you're as you're looking at this map that we now have some more clarity on where the hail risk is, what were some of the areas that really jumped out at you as to be surprising, maybe a higher level concern. 

Andreas Muehlbauer  
So no big surprises here. We always knew that Europe had a hail problem. In fact, I grew up in Munich, Germany, and I remember in 1984 we had a very huge hailstorm in the area that damaged hundreds of aircrafts and Munich airport, causing billions of euros in losses. So it was Deutschmark at that point. Right. 
But so we know that was hail and historically right. And it also was a problem in Europe. What was interesting is to kind of see those areas also reflected in our global hail map. So areas around southern Germany's areas around southeastern France. Northern Italy right is a contender for large hail. 
In fact, the northern Italy area had broken the European hail records twice in 2003, having the largest ever occurred in the in that area and the other region. That kind of was a little bit surprising, maybe is the southeastern areas of Spain, including the area around Valencia where we had a very big flood storm in 2023. 
So and these are the same weather systems that could also generate large hail outside of Europe. I think it's also Asia. Asia has some interesting areas with high risk. In particular, China might be a little bit of a surprise to some people. But according to the data and the records, China has among the largest hail occurrences in the world. 
So not just for that area, but in the world, Japan had some very severe losses in the last few years already. So that's an area on the horizon. That's maybe more from a science perspective. Uh, historically, we always used to think that hail is more a problem of mid-latitude regions. 
So typically areas around 30 to 60 degrees north and south, right where weather systems live. But we also discovered, through our collaboration with the Vietnam Institute of Meteorology and, and hydrology and climate change, that even tropical areas like northern Vietnam, for example, can have damaging hail. 
It's just not as frequent. 

Brian Amaral 
And Anthony, how is this going to help understanding the hazard a little bit more clearly? How, how is that going to help filter down through to the work that the engineers do at FM? 

Anthony Longabard  
So, previously in places like Europe, where we didn't have maps, I mean, we wouldn't have had any sort of recommendations associated with, you know, installing products that are hail resistant. Now that we have, you know, this map, it gives us the scientific framework needed to actually recommend products that are impact resistant. 

Brian Amaral 
By understanding the risk more clearly and more locally in Europe and other places, it could help those places understand and be able to implement things and get products that could actually guard against the risk? 

Anthony Longabard  
That's correct. One of the big ones is roof systems. But then it's also things like skylights. It's photovoltaic panels. Um, it's heat and smoke vents. But then beyond that, we can provide them recommendations. “Hey, listen, you have this high value storage sitting outside. Maybe you should consider moving it inside or providing some sort of cover for it, so it doesn't get damaged during a hailstorm.” 

Brian Amaral 
Cars, aircraft, that sort of things. 

Anthony Longabard  
Yeah, absolutely. I mean, all that stuff will get damaged, you know, in a, in a hail, a major hail event. 

Brian Amaral 
How is climate change affecting hail and the storms that are forming? The frequency and severity of storms? 

Andreas Muehlbauer 
Yeah, that, that's a very good question. And it's, it's an ongoing debate. And there's, there's—it's kind of a mixed picture overall. And that is because there's competing factors. We have obviously the planet warming. Um, that creates higher temperature. It creates the ability for the air to hold more moisture. And moisture is always the fuel of any sort of extreme weather to start with, right? So, if that happens, then in theory, um, the storms that produce hail should also become more powerful, right? More vigorous. And that, that's also what we've seen. But on the flip side, um, that's also increasing temperatures, increasing the melting potential of falling hailstones, right. So, in theory, that should actually make the hail a little bit smaller than it used to be, right. So that's, that's kind of the offset. You have more vigorous storms, but you have more melting of hailstones. And on the, at the same time, you also have wind shear, which is a very controlling factors for the very severe thunderstorms. And that's also decreasing. So, there's, there's different factors at play here. The net effect is a little bit unclear. You know, which is dominating right? Um, which basically means that you get a mixed picture. There's going to be some areas with increasing hail and other, other areas with decreasing hail. And our job is really to kind of find out where is the risk increasing and where is the risk is decreasing in the future. 

Brian Amaral 
And hence the map. Right. I mean, that's the whole idea behind the map. 

Andreas Muehlbauer 
Yeah. The map is, is just the starting point. The next step is going to be to identify the trends and also kind of start developing future looking views to get an understanding. Okay, how is the hail hazard really trending in the future? How is the hail hazard going to be in, let's say, 2050? 

Brian Amaral 
Last question I want to ask both of you. You know, this, this hail map was recently released. It's now worldwide. It uses not just frequency but also severity. So, that's obviously a really important step in, in understanding this hazard. But, but Anthony what's, what's next for hail for your understanding of hail, for the, for the industry's understanding of hail? What's the next challenge that you'd like to see tackled? 

Anthony Longabard  
Uh, the next, I mean, the next challenge I'd like to see tackled may perhaps be dealing with the potential for protecting at an even higher level of severity. I mean, our maps are based on a 15-year MRI. 

Brian Amaral 
MRI? What's an MRI? 

Anthony Longabard  
It's a mean recurrence interval. It's basically a fancy way of explaining the probability of occurrence in any given year. 

Brian Amaral 
Okay. 

Anthony Longabard  
So we use longer MRIs for some other hazards because the consequences associated with those can be very large, you know. Like we said, collapse. Hail, we don’t, we're not dealing with collapse. We're dealing with more aesthetic type damage. So, we use a shorter MRI because the consequence of damage isn't quite as severe. Think about a typical industrial facility. You're looking at a 50-year lifetime. So, you use, you know, a mean recurrence interval right around the lifespan. The lifespan of that stuff that's going to get hit by hail, like roof systems, is much shorter. So, it doesn't make a whole lot of sense if we ask people to use a very long MRI for materials that aren't going to be around that long, you know. Think about a roof. It's going to be around for, you know, 15, 20, 25 years. 

Brian Amaral 
And so, in the future, you're hoping to see additional, maybe, products? 

Anthony Longabard  
You may have, you know, certain clients, certain people that have a really sensitive occupancy, where damage associated with one of those storms would be really, really catastrophic. To me, it makes sense to, you know, if we have a client that wants something, wants a higher level of protection, we have the ability to provide it for them. 

Brian Amaral 
Andreas, what, what are you hoping to see? What's the next challenge as it relates to hail that, that you think is on the horizon? 

Andreas Muehlbauer 
Yeah, the next challenges are really going to be threefold. Uh, number one is now that we have a solid baseline for our, our understanding of what the hail risk really is worldwide, it's now updating this information whenever we have new data coming in. Because we already know it's changing, right? So, the map that we have now, right, won't be the true map in 20 years from now. So, since we, we understand this already, right, we have to kind of find good ways how we can incorporate new data into, into the map and, like, really stay on top of the risk and update this information from, from time to time. Then, secondly, it's also understanding the trends, right? Where does the, the changes in the climate really impact the trends in hail either going up or going down, right? And really, truly understanding it and be able to quantify where it's, where it's going up and where it's going down. And then thirdly, really developing a forward-looking view that we can say, right, the hail risk nowadays is X and in the future, it's going to be plus or minus, right. And be able to quantify this out into the future by 2050 and beyond. Because a lot of, a lot of our clients, they are building new locations, right. And they want to know not just what the hazard is nowadays and how they can protect, but also how that level of protection is going to help them in the future. 

Brian Amaral 
Yeah. And just a quick note for our listeners. We will have a link to the hail map in the show notes, but you can check it out on FM.com. It's in the NatHaz Toolkit. You can check on your own location and see where the hail risk, where you are, what sort of risk you are facing. Anthony, Andreas, thanks for joining the show. 

Anthony Longabard  
Thank you very much. 

Andreas Muehlbauer 
Thanks, Brian. Thanks, Brian. Thanks for having us.