Community Solutions & Market Shifts: Wildfire Risk with EPIC's Patrick Gallagher & Mike McNulty

In the wake of the devastating wildfires in Los Angeles, Season 2 of Building Potential focuses on wildfire risk. These episodes explore what recent events reveal, what the industry still needs to learn, and how we can apply these lessons to build a more resilient insurance ecosystem—and a more resilient society.

In Episode 2 of the Building Potential – Wildfires Special, Patrick Gallagher, Chief Growth Officer of Risk Management at EPIC Brokers, and Mike McNulty, Leader of EPIC’s Winery and Beverage Practice, join Archipelago’s Founder & Chairman, Hemant Shah, to discuss:

• Rising market losses – Why the LA fires could result in over $40B in insured and uninsured losses, with broad implications for pricing, capacity, and recovery timelines.
• The power of collective mitigation – How Napa wineries are collaborating across properties to create zones of resilience and unlock coverage from wildfire-wary insurers.
• Limits of wildfire modeling – Why many underwriters remain skeptical of current models—and how community-level data could redefine how wildfire risk is measured and managed.

Watch, listen, or read along to the full episode below:

Hemant:
Welcome to season two of Building Potential where we're continuing to explore emerging challenges and new opportunities across the property risk management and insurance ecosystem. I'm Hemant Shah, your host, and given the devastation recently in Los Angeles, we're kicking off this season with a series of important conversations about wildfire risk. In our last episode, we discussed the drivers of risk in their mitigation from an engineer's perspective. Today I'm joined by Patrick Gallagher and Mike McNulty of EPIC Brokers to discuss the market's perspective. Patrick is Chief Growth Officer of EPIC Brokers and Mike leads EPIC's market-leading Winery and Beverage Practice. Patrick, Mike, thanks so much for participating in this important conversation and I appreciate you devoting the time to this topic.

Mike:
Absolutely, 

Patrick:
Thank you. 

Mike:
Yeah, thanks, Hemant.

Hemant:
Good. It was good to see you both last week respectively in San Francisco and New York. And while it seems like a long time ago, it was just back in November when you both hosted EPIC's Wildfire Summit up in Sonoma County in northern California, that proved to be a pretty topical and timely event, wasn't it?

Mike:
Yeah, very. And I think as we discussed during and after, it seems that our Galway EPIC events tend to bring out the rain in northern California. So sort of ironic to be having a wildfire event discussion while we were getting six to eight inches of rain on a daily basis in Santa Rosa.

Hemant:
Well, it's one of the beauties of California. It's a wondrous state, including it exposes every hazard known mankind, but you are certainly on brand with the event, given the winery practice and the wildfire we met in the wildland urban interface up in wine country. And the format of the event I thought was quite innovative and compelling as well where you brought together not just a bunch of presenters, but it was really a discussion, a group format to talk about the challenges of wildfire risk and potential solutions across the entire value chain. I don't see that too much in the market. I think the industry often spends a lot of time talking to itself. It was good to see you bring together every stakeholder across the value chain to have an important conversation on this topic.

Patrick:
Yeah, I think that our organization has taken it almost as a philanthropic venture in order to solve the problem of insurance incapacity and affordability, and particularly in northern California, particularly among our wine industry clients. The wildfires 2017 to 2020 created a major business problem for this industry, and we think that there might've been an overreaction by the insurance carriers and they think differently. So we've had a series of events where we've brought them together to argue that case in a more holistic way, introduced them to their clients in a different venue and offered 'em some of the tools to address that challenge.

Hemant:
Yeah, I really appreciated personally the opportunity to participate in that discussion as well. It was really refreshing to see an open-minded conversation between EPIC as in terms of bringing the parties together, your customers and Mike and the winery practice and key market leaders on the insurer side come together to discuss these issues, debate some of the drivers, discuss how we can create more solutions and then create win-win wins across the value chain. I'm looking forward to double-clicking on this a bit later in the conversation, but maybe first, Patrick, the LA fires have happened since we all gathered at your summit. The losses are quite devastating. What do you make of the market scale of the market losses? I know it's a moving target and there's a lot of uncertainty, but what's your current assessment of the scale of this loss from an insurance standpoint and its potential implications across the industry?

Patrick:
Yeah, I think it's important to have an estimate there. This is one way that we can bring together all of the global ca events and add them up. And we've got pretty good data, particularly in the last month because of the fourth quarter financials reporting for a lot of the major insurers and reinsurers. And the current tally is at about 30 billion, and there's good reasons to think that's significantly understated. The first is that the way that contents are being handled on recoveries to avoid the itemization, which is very difficult when you have a total loss, you didn't think to write down all of the things you own in your life. And so there's been a soft regulatory push and lurking in the background is a harder legislative push to ensure that insurers essentially overpay for some of the claims here, including up to 100% of the contents limit, which is really usually unlikely in most of the homeowners policies that we place. So there's actually not much in contents there,

Hemant:
And many of those homes had significant contents limits and significant contents exposure. So it's not just the structures that we visibly can see in the photographs, but it's also the contents and the additional expenses can really add up on some of these higher-value homes

Patrick:
For sure. And the other thing is when we think about catastrophes and we say, oh, hurricane Michael was 25 billion, the real number that we should be concerned about is a society is the total economic loss, but it's a lot easier for us to tabulate the much, much smaller insurable loss. And I think that that's one of the reasons that this 30 billion might be significantly understated is because of the huge uninsured exposure and the recourse for these people is to go after, at least for the Eaton fire, the power companies who potentially started the fire and they have significant claims-paying ability up to 20 billion through a collective insurance scheme. And I think that that could be a 10 billion loss to add to it and get you up somewhere in the 40 billion range.

Hemant:
So you're thinking 30 billion plus plus could be the ultimate scale of the losses.

Patrick:
Right? And then there's a number that'll never be tabulating, which because this fire was so big, a lot of it got into the layers of the insurance-linked securities markets and retrocession markets, which could potentially erode an aggregate limit really early in a policy with a lot of runway and wind season to go. And if those aggregates hit, it's not really clear how you tally that and we won't know maybe for a year's time, but that's just another number lurking in the background to get it over 40.

Hemant:
Yeah, compared to casualty lines, property is certainly a short-tail exposure, but in these large complex catastrophes, as you note, it can take some time for the full extent of the loss to develop across all the different vectors of claim and recovery and settlement. So it could be you're flagging, it could be a year or two before we know, but it could clearly exceed 40 billion. I think one other dynamic, Patrick, is that I'd love to get your perspective on that makes this maybe a little different than some catastrophes. Is that – at the heart of some of the coverages, the California a fair plan? And so the dynamics there, maybe just remind our listeners how that might play out in the marketplace.

Patrick:
So the fair plan's, claims-paying ability is not fully defined and hasn't been fully tested yet, but essentially the fair plan, like the fair plans in many states offers a skinny down version of coverage with limited coverage overall. So the full value of the home isn't covered and that the contents and some causes of loss won't be, there's no liability coverage, et cetera. What they do have in addition to a reinsurance program is the ability to tax other insurance companies and who then in turn have the ability to, tax might be the wrong word, but to assess, yeah. Who have the ability to upcharge and assess on a portion of that of their own policies. California is a very activist insurance regulator and has partial control over the activities of the fair plan and the insistence on paying claims to the fair legal contractual level or above that. So it's not clear exactly how that'll pay out, but what they're reporting right now is that they're going to use up all of their reinsurance and then some as they assess the competing insurers for some of the losses.

Hemant:
Earlier you mentioned that there's been a number of earnings calls, which have shed some light on the extent of the losses. I listened in on one of those calls from the CEO of AIG and Peter Zaffino flagged that while it's just the start of the year, given the scale of the LA fires 2025 could easily exceed 200 billion in insured loss. When all is said and done, what do you see the impacts of those kinds of scale of losses on the market itself? I mean the market has, property market has been relatively speaking, softening a bit the last couple of cycles. Do you think these kinds of losses that we're seeing in LA and then the potential for losses through the rest of the year could change some of the market dynamics out there?

Patrick:
It could be $200 billion is a big number and it would probably just about tie in real 2020 $5 the worst cat year of all time with 2017. And I think that we'd be able to weather the storm, but certainly 200 billion would reverse the stabilization and softening of the property insurance market globally. A IG, like any carrier right now who's heavily exposed in California has a reason to make a little bit more of a meal out of this than warranted. And I think the only indications of hurricane seasons are that it's likely to be a mild one after significantly elevated in the prior two years with some cooling of the waters in the Atlantic. So I think he might've been assuming that the near historical trend of hurricanes would continue in that number, but it is early on, so I take his point, but my guess I take the under 200,

Hemant:
Well, we're here in March and we've already got 30, 40, 50 billion of insured losses on the market scoreboard, and there's still a long calendar in front of us in 2025. Mike, clearly LA has significantly exposed and damaged homes. I think I read somewhere that estimate is 85 to 90% of the insured losses are likely to be from personal lines policies. But I think firsthand that fire risk is also a concern in commercial lines. And your customers in particular, some of them have significant exposures to wildfire hazards. What's it been like securing coverage and working in the market with your customers? Because it's not just the LA fires, it's we run the clock 5, 6, 7, 8 years. The tubs fire in 2017, the campfire in 2018, the lightning complex fires in Northern California in 2020, Lena Maui in 2023 and now LA 2025. What's it like being from the perspective of commercial lines, customers of yours looking to get capacity for their wildfire risks?

Mike:
Well, it's been extremely challenging for all the reasons you mentioned, and wildfire remains a tough proposition for many of our winery clients. But one thing that the excuse did sort of bring to bear is that since 2017, Napa County, I'd say in particular has taken a very aggressive approach to what I would call countywide effort to really get a handle on their wildfire exposure and address it in a very proactive way. And what that's doing is creating a lot of enhanced resiliency sites among the various wineries in specific areas. For instance, Prichard Hill, which is on sort of the east side of the valley. So what we've got is a very tough market that has awakened all the winery owners in that area particularly, and they've really banded together. They're really focusing on mitigation of wildfire, anything that could potentially enhance the ability of a wildfire to take off the dead trees, the forest floor covered in dried brush and things like that.

So the enhanced resiliency of the entire Napa County is really helping us get through and get more capacity out of carriers. We're seeing that more and more as the winery owners themselves go from 10 years ago in 2015 to winery owner could go out and have six or seven admitted carriers in California, their winery account, property general liability out of the whole thing. Very simple and straightforward. Now it's much more difficult. We're piecing together large layered programs for property and many of the standard carriers that you would've remembered from those days are no longer writing insurance for wineries in California. So it is been very challenging. I say that the winery owners, I think they're naturally used to working together. They're not the kind of competitors that want to eliminate each other share. No,

Hemant:
It's a tight-knit community of like-minded people.

Mike:
Exactly. And they understand too that wildfires don't, they don't know that they're crossing over from one winery's area to another, so the wildfire is going where it's going. So it's in everyone's interest to really get behind this effort to mitigate risk. And that's, I think they're really pooling resources. They're working together and I think it's causing the insurance industry which is trying to catch up. And in some cases they do rely on their somewhat outmoded modeling systems that we need to bring to bear and bring up to date, as you know. So there's a lot going on. It's a moving target. I think there are best in class. There are certain wineries that are doing more than others, and we're trying to sort of lift everyone up and get that awareness to where it needs to be.

Hemant:
Well, you raised a really important point, which is that wildfire risk perhaps. Unlike some perils, you can bend the curve with mitigation
And not just every man for himself mitigation, but community level resiliency. And I was struck at your summit meeting last November, the emphasis not just on localized mitigation, which is effective, but how as you say, do you form a zone of resiliency among adjacent owners who have a mutual interest in ensuring the whole region doesn't have a fire? And you can create a kind of herd immunity by having multiple parties come together in a fire zone to all take collective action to watch each other's back. And I think that's something that I'd like to come back to because I think it's an important dimension of fire risk that may be not is well recognized by the modeling community and by some of the markets who tend to have a mindset. I think it was at your symposium, one of the lead underwriters had an aha moment when we started dialing in this conversation of community-level resiliency where you're trained as an underwriter to avoid concentrations. And Joe, you want to place multiple diverse bets, but with wildfire, you might be better off playing a bigger bet on a zone of resiliency where it may be geographically concentrated, but you've created a zone of resiliency because everybody's mutually reinforcing everybody else's mitigation with defensible space, controlled burns, defensible perimeters and resilience structure. So I think you flag an important dynamic that is emerging out of the recent events that can drive more and therefore more capacity.

Mike:
Absolutely. No, it's been really interesting to see the development of the effort. And I think if someone wanted to dive a little deeper into this, there's various wildfire councils, they're called firewise councils throughout NAPA and Sonoma and probably coming to be throughout California at some point that are really taking the lead on getting together the grants that are available from state and local governments and federal government and then matching those up with individual and groups of winery owners and doing these resilience efforts as a group. And so it's impressive to watch. It's making a big difference.

Hemant:
Yeah, I have some affinity for this because we spend a lot of time personally up in Sonoma County. We have property up there and after the Tubbs fire in 2017, my own community came together and got some grant money and took some collective action, multiple property owners along a common road to not only improve the risk profile each of our individual homes, but create a zone of resiliency that actually led to a more defensible area that led to good outcomes in the lightning complex fires in 2020. So I've seen it work firsthand.

Mike:
It does work.

Hemant:
Patrick, Mike mentioned modeling in there and some of the challenges of modeling. I think many in the market take the view that when it comes to catastrophe risk, its models make markets, and when models are challenging to build or to understand the risk, it's hard to make a market if you don't have confidence in the underlying ability to quantify the risk. What do you make of that dynamic of models making markets wildfire risk, ability to model it and some of the challenges and scaling capacity for this peril?

Patrick:
Yeah, I think I broadly agree with it maybe to modify that widespread institutional acceptance of models is what make markets so good point. The stock option market was black, sho, everybody agreed with it, everybody went with it. Obviously the hurricane models, it wasn't enough to everybody say, oh, this is really cool and helpful. We had to speak with the folks in Florida to make sure that you understood it and allowed it in the underwriting. And these wildfire models I think are particularly challenged by underwriters. They have a natural skepticism in part because there isn't a great catalog of historical claims. We do have some and we do have insurance, a catalog of insurance events and a catalog of actual fires. But it's kind of like cyber insurance where who cares what kind of cyber attack happened 10 years ago? What relevance does that have at all because of the changing technology? I think that's kind of the attitude. The wildfire models are incredibly intricate at their base levels. They start with a map and they say, you can see there's a lot of trees over there. There's a lot of wildlife, urban interface very close, and the inability of responders to contain a fire there. And so that's a high risk. But most of the models have taken a step further in two areas. One is to think about the likely path of the fire, which locations would be more exposed than others because of the topography and the nature of the terrain. And then a second level is to think about the defensibility at the home level. And there's a lot of data that you can throw at this problem. Not all of it's structured, very little of it's being collected in a systematic way.

And so there's the right parts there for a super sophisticated model, but most of the models that we've seen so far have focused either at the footprint level, where's the fire likely to be at the micro level to say, yeah, we know it's coming to your neighborhood, but will your house ignite or will it pass your house? And then at the intermediary level, which is of particular concern for the types of initiatives that Mike's talking about, where we're saying, well, we're just going to look at these 50 acres here and make sure that that's a safe place and that the firemen can get in and do their job and will be aware well in advance that there's going to be a fire there and we'll have water on site. And so there's models for all of these and there isn't yet a model of everything and the underwriters might be not investing of in creating all that just because of snap judgements about the ineffectiveness of the historical cattle.

Hemant:
Yeah. One of the things that strikes me as a recovering catastrophe modeler is it feels when it comes to wildfire, we've rewound the clock about 20 years where there's methods and emerging models for different elements of the risk, but that a lot of the markets remain quite skeptical and as a result, take it almost accumulation based approach to managing the risk where they're managing accumulations in zones as opposed to putting a lot of stock on the ability to diversify, measure the risk profile to the point where you can build a more significant book of business based on certain bets you're making that the fire's not going to burn every single structure down in this area, I think, but there are emerging as you flag, I believe you've been studying this space for some time to help your customers better understand how they can manage their risk and the markets understand how they can better allocate capacity to this risk.

Patrick:
We have, we've been looking at all of the vendors and there's interesting things particularly in the area of computer vision, taking a picture from above, but also at ground level of a house and saying, Hey, are those the right shutters to have? Are those the right vents? How far is that tree? Are those three trees in conductivity? So using these algorithms based on these digital images have probably three distinct solutions. There's also been a lot better satellite image ring, which is more common. So Google has a big initiative to catch fires at an early stage. There's been autonomous technology both in form of flying drones, but also cameras set out in the wilderness to try to interpret what they see as whether it's the start of a wildfire or not. So there's all this information that isn't quite linked up yet. And I think the one that I find most interesting and exciting is what's called ing to pause for a second, and because it's called the Wooey data
Commons. Excuse me again. So I think what I find maybe the most interesting of the technological advances is at a nascent stage, the WWE data commons, which is a project to collect the information about individual homes. And I hope also to collect and organize some of the data around community resiliency. The best way to abstract it right now we have for those community efforts is how much money is being put into it. And we can think of an ROI, even if it's a negative one, to at least to measure the relative benefit of all the work that's being done in Northern California and that spread.

Hemant:
Yeah, the traditional modeling approach is still relevant for fire. You try to bottle, as you say, an ignition, a vector of spread given vegetation, climate, precipitation, humidity, wind, and then the localized vulnerability of the structure and the media area around the structure. But you also flag that there's a couple of additional dynamics which are challenging. One is how do you represent the community aspect of the resiliency where it's not just an individual home or structure with its characteristics, but how a community has come together to create a zone of resiliency and getting that kind of insight from the we data commons from the firewise initiatives into the modeling. Because if it's not in the modeling, it's hard for the underwriters to give credit to that. It's one thing to intellectually understand it and to talk about it, but you have to quantify it. And the other feedback loop, which is challenging that we discussed at your summit is the fire response itself.

Unlike many or almost all other perils other than cyber, there's a response vector. And the extent of the loss at a property, a community or a region is also a function of what's the fire response effort, how many tankers can get in the air, how much concentration of bulldozers and crews and firefighters can be deployed to a particular zone of a defense. And these feedback loops also are influenced by mitigation because the fire responders have to make tough choices as we know from a firsthand experience, they have to triage their response, and they're more likely to devote resources to defending property that can be defended. So the more you mitigate and the more you build community resiliency, not only is that a priority quite effective, but it draws in responders to defend your property, which means you won't have a loss. So there's a feedback loop between the mitigation and resiliency effort and the fire response effort, which is also challenging to model, but it's important. And as you say, if you can't quantify it in a standard way, it's hard for the market to give credit to that. But I think that's an emerging frontier of capability that needs to be quantified and promulgated throughout the marketplace.

Patrick:
And I think you spoke poignantly at our event about being on the right side of that triage at your own home, and it seems a super hard modeling problem that you've also introduced to say, well, how do we decide which of these is near the firehouse is an easier way of looking at that with traditional fire. But when it comes to wildfire, what kind of priorities that maybe the local fire authorities haven't even decided in their head yet? Let's try to anticipate what they'll do is they're really nice.

Hemant:
Well, we all know that all models are wrong, some are useful. So it's time for the useful models to emerge. And as you also noted Patrick, it's not just important for the model to be effective, but it needs to be adopted across the marketplace to create capacity for this peril. And I think we're at the cusp of that as we wrap up. Maybe we can return to what motivated this conversation in the first place, which is the LA fires. I mean, it's only been a few weeks and the city is, and those affected neighborhoods that Palisades now tna are just starting in the long hard process of recovery. What insights do you guys have from previous experience and from your market perspective on what the city and the stakeholders in the region need to be mindful of to build back better and in a way that enables more scalable insurance solutions to be delivered? Because this problem is not going away, and if we don't come up with scalable solutions across the value chain, there's going to be real dislocations not only in the insurance market, but in our society because insurance is in a key enable of a lot of economic activity and social activity. And if we can't solve this problem, we've got a real challenge on our hands.

Patrick:
Well, Hemant, I think that we've got a community problem and we need a community solution. For many people, the catastrophe is ongoing who haven't returned to their home or will not for over a year. And that's their schools, their jobs, their communities. And we've got an issue. It's kind of exemplified the California and some people's minds around how we build in this country.

And the zoning and the permitting aspects of this are often a very regimented way of doing things. They're good ideas, but they're all on an individual basis. They're making judgments. And here's an opportunity because we have these rules of resiliency, but especially because we have affordability and the speed of making all of this rebuilding happen simultaneously, that I think in the same way that maybe the insurance department is bending the rules a bit about how policies are recovered, it would be great if we could also bend the rules to reflect our competing values, but prioritize as we often do in an emergency, the victims and what's in their interest right away.

Hemant:
Yeah, this is a community problem and needs community-level solutions, and there needs to be a balance between the importance of getting people back in their homes and getting on with their lives so they can rebuild and recover and to take measures to ensure that we're actually building back better and not just setting the stage for the next disaster. And I think managing that tension, because it takes time, unfortunately, the real world, it takes time to understand the key lessons learned to do the hard policy work to affect change. In the meantime, people want to get back in their homes, back on their property, back with their lives, and managing that tension. We can see it playing out in Los Angeles already between let's just get people back versus let's take a beat and think about how we want to build back better this time around and what that really means and what are the trade-offs we need to make to ensure that we're building more resilient communities as we recover, not just building back the way it was before.

Mike:
Absolutely. Very true. I think it's a chance too for the Southern California communities to adopt that firewise mentality where there's a concerted effort within a neighborhood or a community that reflects all of the things you were just mentioning and puts them into play as the rebuilding process begins so people really understand what the correct materials are to be used for their home, how close they should be planting shrubs around their house, and then also what's going on in the general community, what you are adjacent to a national park. We need to keep an eye on that. How do we protect ourselves from that? LA has pretty predictable wind every year. They know when that's coming. You've got to be ready and you've got to take that community approach and really take all those challenges head on. And I think that would be a great takeaway from the tragedy that we saw in January.

Hemant:
I think that's a great note to end on. And the theme of wildfire risk as being an interconnected challenge and an interconnected opportunity modeling it is very interconnected. Mitigating it as interconnected and recovering from it is interconnected. -Thank you, Mike. Thank you, Patrick, for flagging all three levels of interconnection that we need to be mindful of as we look to quantify the risk, mitigate the risk, and recover from this risk. And that's an important call to action. So thank you both for chatting with me about this topic, and I look forward to continuing the conversation with you guys offline as you search for new ways of thinking and new ways of modeling and new ways of enabling your customers to access capacity for the risk and do so in a way that they also demonstrating their ability to mitigate the risk, which creating the positive feedback loop for all parties of the value chain. So thank you for your time and look forward to continuing the conversation.

Mike:
Absolutely. Thanks.

Hemant:
Thanks, Mike. Thanks, Patrick.