Speaker 1:
From the New York Stock Exchange at the corner of Wall and Broad Streets in New York City, welcome Inside the ICE House. Our podcast from Intercontinental Exchange is your go-to for the latest on markets, leadership, vision, and business. For over 230 years, the NYSE has been the beating heart of global growth. Each week, we bring you inspiring stories of innovators, job creators, and the movers and shakers of capitalism here at the NYSE and ICE's exchanges around the world.
Now, let's go Inside the ICE House. Here's your host, Lance Glinn.
Lance Glinn:
Meteorology plays a vital role in financial markets, shaping industries such as agriculture, energy, insurance, and commodities trading. Accurate weather forecasts help businesses manage risk, whether it's farmers adjusting crop strategies or energy companies preparing for fluctuations in supply and demand. Extreme weather events can drive market volatility, influencing everything from stock prices to commodity futures.
Joining us today is Dave Margolin, Meteorologist for Intercontinental Exchange. That's NYSE ticker symbol ICE. With over two decades of experience in the field, Dave began his career at the Weather Channel in 2000 and has since held roles at AccuWeather, Reliant Energy, and Citadel, among others. For the past four years, he has been leveraging his experiences and expertise in his current role at ICE.
Dave, thanks so much for joining us Inside the ICE House.
Dave Margolin:
Yeah, it's great to be here. Thank you.
Lance Glinn:
So when I joined Intercontinental Exchange, I was frankly surprised to learn that the company had a meteorologist. My first thought was, "What impact does weather possibly have on a company like ICE?" And as I delved deeper and I became more experienced with the company, I came to understand its significance. But for those unfamiliar who were like me when I first started, how does weather influence ICE, its core businesses, and just financial markets overall?
Dave Margolin:
Yeah, and that's a very fair question. I've gotten that question quite a bit over the years, and especially for folks who are not in the financial or energy sector, but it turns out that weather and climate really have a significant impact on the economy. I think the latest numbers I saw was weather can impact or change the GDP by as much as 5% each year, so that's over a trillion dollars. And of course, you think of a company like ICE, very much tied to the economy through the exchanges, through our data services, through the mortgage, so you see that connection there.
And more specifically, when you think about energy, I think the relationship is even stronger because what do people do in the winter in terms of their thermostat, their heating bill, or the heating in the house? You turn the heat up when it gets really cold, and when it's a little bit warmer, maybe you turn it down and such. And that phenomena is going on all over. And so what they're trying to do, what the energy community in particular has done is realized, "Hey, listen, if we can get weather information and weather guidance days, weeks, or maybe even months in advance, there's a lot of opportunity for us and ways that we can mitigate our risk."
Lance Glinn:
And your research, your valuations, what you do on a daily basis, they enable you to obviously closely monitor incoming weather events. We've had a high frequency of weather events over the last few years. But when it comes to organizations like ICE, how can they, along with traders and investors, leverage accurate meteorological data to gain a competitive advantage in markets?
Dave Margolin:
Yeah, and ICE, As I think you alluded to, ICE does have a lot of this weather data in house, but certainly other groups do too, so there's really a wealth of information now that's out there. We've got, and a lot of this is publicly available, especially in the United States, there's weather models now that go out hours, days, weeks, and even months into the future. There's data, historical data that goes back 30, 40, 50 years, and a lot of that data can be used for analysis in terms of understanding relationships.
So the wealth of data, especially, I would say, over the last 10, 15 years with cloud computing and just data becoming so much cheaper, has become much more readily available. And in part, that's actually why you've seen this explosion in weather companies, startup weather companies, over the last 10 years. 20 years ago, to be able to source that data was incredibly expensive, and so only the major, major companies or major modeling centers could do it, but now a lot of startup companies have had access to this and that's been a great thing. You've seen an explosion in value-add weather services and products.
Lance Glinn:
So Dave, take us back to the beginning because I think we all, at some point of our lives, specifically in our youth, have envisioned, "Man, it'd be so cool to be a weatherman." I remember going on a school trip, I don't remember what grade I was in, but we went to a TV studio on a school trip and they showed us how to use a green screen. And that obviously conjured up visions and dreams of, hey, being in front of that green screen, telling the weather on a daily basis.
Is that what got you interested in meteorology, this potential goal, this potential dream, desire to be in front of the camera, be a weatherman? How did your interest and how did your career in weather really start?
Dave Margolin:
Yeah, you know, it's a peculiar thing, and I'm not going to embarrass myself too much by telling what other kids thought of me back when I was younger, but from as young as I could remember, I just loved weather. I don't know where I got it from. I didn't necessarily get it from my parents. They weren't anti-weather, but they weren't necessarily meteorologists or scientists, but I had this fascination and love for weather, in particular snowstorms.
Lance Glinn:
So were snow days your heaven?
Dave Margolin:
Oh, of course, of course, but like any kid. But I would be the kid who would be watching the Weather Channel and all the weather services. We'd have sleepovers and I'd be like, "Hey, guys, can we just turn to the Weather Channel or something like that real quick?" Or if there was a chance for snow, I'd be like, "Hey, guys, I've just got to go outside and check what the weather is."
So I don't know where that came from, but it was there, such that by the time I was at a point to look at colleges and things like that, I was like, "Well, maybe I can make a career of it. I don't know. But I just know I like it," and I went into it, enjoyed it, and here we are.
Lance Glinn:
And how did the specific role, because again, when I joined ICE, I didn't anticipate the company having a meteorologist, so how did the role, not necessarily with ICE, but just in general in the greater financial community, how did the role of a meteorologist come to be?
Dave Margolin:
Yeah, so I can't speak to the day one necessarily, but there are definitely some early stories that I'm aware of where that connection was made, and Enron obviously had a bad end to it, but Enron was one of the first companies that really realized the value of meteorologists. They actually had, I believe, two or three teams. And it's an interesting story within that too because I believe the president or CEO was Jeff Skilling, and his brother was or is Tom Skilling, and Tom Skilling is a very famous meteorologist in Chicago. I think he just retired recently. So there was that connection there, I believe, where Jeff knew from his brother how important weather is and how it can have such a big impact and such.
And so Enron was one of the first companies that really leveraged it, especially for their energy trading business as we had talked about earlier with natural gas and oil. And obviously, even though Enron is no more, it collapsed, other companies took note of that and started hiring meteorologists. And of course with ICE, it makes sense. Again, as we talked about earlier, ICE, a major financial company, has a lot of people who use their products to trade natural gas and oil and other things too, and so why not have their platform, already great with the financial information, why not even make it better with the weather data?
Lance Glinn:
So I'm sure millions of people are out there like me. Every morning, I get up from bed and I check the weather. I'm coming from New Jersey. I want to see, do I need a jacket? Am I okay without one? Is it going to rain? Is it going to snow? What do I need to wear to make sure that my commute from home to the New York Stock Exchange and back is comfortable?
Temperature is pretty easy to interpret, right? You'll see 30 degrees, you'll see 35 degrees, you'll see 40, you'll see 80 when it's in the summertime. And you know, just by looking at the number, "Okay, it's going to be warm, it's going to be hot." But there, of course, is more complex data, like you were talking about with percentages of rain, with percentages of humidity that I'll see, especially the humidity part, and I'm like, "Well, what does that mean to me? How do I interpret that data?"
So how can meteorologists, and this goes back to your point of communication, but how can meteorologists present the information in a way that's easier for consumers to understand and helps them make better decisions, make better-informed decisions like I have to when I'm getting up at six o'clock in the morning to obviously come here to the NYSE?
Dave Margolin:
Yeah. So again, that's a difficult question. These are certainly challenges that we have, so I don't have a perfect answer, but just some thoughts on it.
So one is I sometimes think that we underestimate the ability for just people to digest probabilistic information. An example of that is the hurricane center cone, the forecast center, the forecast cone. And everyone knows about that. And that is a great example of a probabilistic forecast. It's been around for a number of years now, and I think most people and the emergency responders and different community leaders, I think they understand it. They know what to do with it. They know, hey, if you're in the cone, there's some risk. If you're at the center of the cone, there's probably a higher risk, et cetera.
But we don't apply those same rules, and as you alluded to a little bit, to temperature and precipitation forecasts often. Especially with temperature forecasts, if you likely pull up Google or just go look on the TV, they'll put up their 10-day forecast and you'll notice it's deterministic, at least most that I've seen, from day one all the way up to day 10. And when I say, "Deterministic," it means they just give you one centerline forecast. Well, we know, most of us, and I think most people at home, of course, know that the accuracy of a day one is not the same as a day 10.
So I think what would be much better is to provide probabilistic forecast, if you want, still provide that centerline, but provide this probabilistic forecast to those users so that on that ten-day forecast, you would see something like a very tight range around day one. Let's say the centerline was 70, I might say, "Between 68 and 72," but by the time you got to day 10, it might have a five or 10-degree range around that.
And where I think that would help, to your other point, I would say more about impact, is, I don't know, if you're planning this weekend trip, and I get this question from my wife all the time when we go out, she's like, "Do I need a jacket? Do I need a sweater?" And this kind of thing. Well, if you know that the range, you're going on a weekend trip up somewhere and the max temperature is likely going to be between, I don't know, 65 and 75 as opposed to just a point, you can pack accordingly. You could say, "Well, there's some risks. I'm going to need a light jacket, a sweater. Let me put at least one in there." But there's also a chance that it's just going to be in the 70s and a couple T-shirts.
So I think that's probably a very simplistic way, but how we can adapt better if we provide better, more probabilistic or information about the risks of the forecast to the community.
Lance Glinn:
Do you think long-term predicting is getting better though? Because like you just said, you get a 10-day forecast. Day one pretty much on the mark because it's obviously the soonest day coming. Day 10 is sometimes a hit, but sometimes it'll say, "Oh, it's going to be sunny, chance of 80-degree temperatures, no cloud in the sky." And then ultimately, you get to day 10 and it's raining and it's 70 degrees and it's a thunderstorm.
So with that being said, do you think we are getting closer or predicting that day 10 is getting better over the last five, 10 years, or do you still think it has a long way to go?
Dave Margolin:
Yeah, I was trying to avoid your question about performance because meteorologists don't like to talk about it. No.
So a couple things. So one, I think overall over the last, again, 10, 20 years, few years, forecast performance overall, whether you're talking about temperatures, precipitation, tropics, has gotten significantly better. There's some numbers, and I don't remember them exactly, but they'll say if you compare a day-five forecast today or the accuracy of a day-five forecast, it compares maybe 20 years ago to day-two forecast. So it's really come up dramatically, but people get used to that and they want more. So they're like, "Okay, I trust day five," but as you were saying, day 10.
In terms of where we're going into the future, I think we're going to continue to make improvements, but there is question about how far and how much as you go further out. And some of that, there was a scientist a long time ago, his name was Ed Lorenz, and he was the one who discovered what later became called the butterfly effect, and a lot of people may have heard of that. That's the idea that a little butterfly flapping its wings in Brazil or something ends up causing a tornado in Kansas or something. But what he discovered and what he found, and this is where one of the great limiting factors of weather forecasting is, and he discovered this by accident, but all these observations, temperatures and such, go into these weather models from all over the globe and then the models make forecasts.
Well, the instruments that measure temperature and such of course aren't perfect. Nothing's perfect. It's created by humans. They're also not standardized. The thermometer we might use in New York City may be very different, a different manufacturer than in someplace in Africa or Tokyo or what have you. But those differences and those errors, actually, even though they may seem trivial and really small, like instead of 23.2, it's 23.4, what he discovered is they compound. And as you go out hours, days, and especially weeks, those errors compound so much that the forecast can actually become such that it's low-skill or no-skill above climatology.
So I think, that's a long-winded answer, but we're going to have to deal with that. Some meteorologists, some scientists don't ever think we'll never completely get past that. So it's always going to be something hard to push, but there's still a lot of opportunities to make incremental improvements for the years to come.
Lance Glinn:
So in the more recent months, dating back to the last few months of 2024, obviously now the first month of 2025, we've seen quite a few of these significant weather events happening throughout the country. Since early January, the Los Angeles area has faced a series of devastating wildfires, destroying homes and destroying communities.
When it comes to forecasting these disasters specifically, whether it be a wildfire, whether it be a hurricane, whether it be an earthquake, whatever it might be, what are the biggest challenges in predicting when and where they'll occur?
Dave Margolin:
Yeah, it's a very difficult situation because we know a lot of people like to live in some of these climates. You think of Florida, the Gulf Coast, a lot of people. There's been a tremendous buildup of people, companies, corporations, business along those areas because it's warm, it's mild, what have you. Same thing for Southern California. Unfortunately though, those areas have significant risks from different types of weather events. In the Gulf of Mexico or Gulf Coast, you have obviously hurricane risks, significant hurricane risks, and as you mentioned in Southern California, you have the fire risk.
So the forecast for those obviously continue to get better. The communication aspect should get better. I will say with this recent event in Southern California, the forecast, the weather forecast, relatively speaking, relative to what's available at this time, were relatively good. In other words, there were the forecast centers in those areas that published the fire forecast, they made forecasts that were projecting a very high risk for fire danger several days in advance. Now, we talked earlier, our communication obviously has to get better, but really, I think that the things that obviously have to be looked at, and I think they're already starting to are mitigation measures.
You're living in an area that's dry, you can't control or change the environment, or at least it's not easy to change the environment in Southern California. But what you can do is with the homes, obviously there are different materials that they can look at and explore that might be safer from fire risk. Vegetation is another big one. Yeah, you might like certain vegetation in your community or your house, but it may be very susceptible to fires. So changing the vegetation, the landscape there can be helpful. And of course, providing protective layers.
So there's some things there that I think they're going to look at, and I think the other thing here too, and certainly ICE could be a player involved here, is looking at how the markets can help, and some of those markets meaning like the insurance markets. It seems like we're starting to see this, especially in the last few years, whether it be hurricanes or I think hopefully with the fire dangers, is that the insurance markets can do a better job in terms of pricing in those risks to make people more aware and the communities more aware, but also to incentivize better to say like, "Hey, those mitigation measures you and I just talked about, if these communities, if these homes put in those mitigation measures, which can make it less likely, we're going to provide you with discounts."
Lance Glinn:
And wildfires, look, they have far-reaching effects on various sectors. Obviously, we talked about the community and the surrounding areas and how wildfires affect them, but you talk about insurance, as you just mentioned, energy sector, agriculture, commodities. How do you factor in potential wildfire threats and actual then wildfire events when analyzing various market trends, risks, and ultimately what you do for ICE?
Dave Margolin:
Yeah, I think that's... I'll piggyback off what we were just talking about with the insurance markets. I think unfortunately, I think there was a disconnect, I think, up until recent years in terms of actually pricing in that information, whether it be about climate or weather risk, hurricanes, fires, heat waves. There was somewhat of a disconnect for various reasons to that. But it seems like the market is really starting to pay attention to that now and try to understand those relationships and bake it in. Mortgage services too. Mortgages is, you're hearing this in the news quite a bit, not just with ICE with other groups too, that they're realizing if they can better assess the climate and weather risk of different areas, whether it be severe weather and tornadoes, hurricanes down here, winter storms, then they can better price to those events.
So I think we are seeing an adaptation from the markets to starting to figure this out, but I also think it's going to be be a long road that's going to have to involve some collaboration between government and the market services.
Lance Glinn:
So we talked about the wildfires happening in the Los Angeles area, Southern California. In the Southeast, they faced Hurricane Milton and Helene in late 2024. And many areas, especially in North Carolina, are still dealing ultimately with the aftermath.
So when it comes to natural disasters, you talked about communication, but with these natural disasters, how can communication continue to be improved, potentially helping to limit impact and then ultimately saving lives of the people that are most affected?
Dave Margolin:
Yeah. So this, again, is a really difficult question, but I think it comes down to education. I mean, I think there's a few things that certainly we can look at, education being one of them. The more that all levels that are involved in this, from the forecasters, the meteorologists, the modelers, all the way down to the homeowners, the community folks, local government officials, I think the better that we communicate and educate, the better off, people are going to make better decisions because ultimately they have to make decisions. A lot of times it comes down to the homeowner to make a decision, and the more informed they are, the better.
So I think that's part of it. And I think creating pods, and we are doing more of this now, but creating pods between all those folks who are involved in the decision-making I think is extremely helpful too. And I think really, in tropics like with the hurricane center, I think they're doing a really good job with this. What I mean by these pods are that you're going to have the meteorologist literally in the same room or the same virtual room, if you will, with government officials at local, state, federal levels, and then also, again, the emergency responders on the ground. Because the more that they're communicating and able to discuss and talk, the less risk there are that you're going to play that operator game and information is not going to get from point A to point B.
So again, I think we are seeing that with some of the disaster weather events, especially with tropics, but I think there's definitely an opportunity for improvement in some of these other areas.
Lance Glinn:
So I do want to touch on climate change real quick. Obviously a widely discussed topic. We're witnessing a rise in these extreme weather events. I remember on a podcast late last year, we talked about, and I don't remember the numbers exactly, but this increase year over year of the frequency of these weather events as well as the intensity of these events too. These go hand in hand.
So to what extent do you believe, broad question here, but to what extent do you believe climate change is driving the frequency and driving the intensity of these wildfires, of these hurricanes, of these natural weather events that have such a enormous impact on obviously the surrounding communities?
Dave Margolin:
Yeah, I think it's a little bit of a nuanced answer to that, and it's a difficult one because you see it in the news and just from various sources that sometimes these things can get so conflated in terms of climate change and what it means in terms of weather. And what I mean by that is it's not the same for every weather phenomena.
And so for example, it's my understanding, based on the consensus of the climate community, that there is a very high probability and high confidence that climate change is having a major impact and will have a significant impact going forward on things like temperature, and specifically heat waves. So there's very strong data that suggests this, so this is absolutely something that if you're vulnerable to this, if you're in an area where heat waves could impact the socioeconomics, it's definitely something that you absolutely need to take very seriously.
Coastal flooding, another one where there's high confidence that waters, especially along coastal waters, coastal areas, think like the Gulf Coast, what have you, are going to continue to rise. Again, something that could have extreme impact on human life, but also the economy too because you think a lot of major cities and infrastructure is built right, like here, right along the coast.
But then there's some other weather phenomena where it's less well understood, and I'll go through the different levels here. The one that's in the middle, a little bit gray where they think there's some relationship but they're less confident and it's less confident exactly how strong it is, is tropical cyclones, especially when you look at global tropical cyclone activity. I won't get too into it, but there's multiple factors that control tropical cyclone activity and how climate change is working.
In some cases, climate change may have an impact where it's certainly going to increase those ingredients or factors that could lead to more, like sea surface temperatures getting warmer. But in some of the other factors that control activity, it may actually work against it. So scientists are a little less, especially when you look at global activity, less confident in terms of exactly where we're going in the next 10, 20 years.
And then all the way on the other side where there's, again, very little data or no data are things like tornadic activity, severe weather. The relationship there, again, is much weaker. Part of it is because severe weather, and especially when you talk about tornadoes, it's on such a small scale and climate change is obviously on a much larger scale that the jury's out or science still needs to understand that.
So the takeaway from that is definitely try to understand, especially if those different weather phenomena, you may be in an area susceptible more to tornadoes verse tropics, definitely get that information because again, climate change is not going to impact all these different weather phenomena the same. It's also not going to impact all areas the same, i.e. like the impact on, say, Canada could be different than, say, the impact on Mexico.
Lance Glinn:
So I want to shift to AI's role in meteorology because obviously it's a technology that's influencing almost every business sector. So for someone in your role who's analyzing data and risk, how is it transforming weather forecasting and improving pattern prediction?
Dave Margolin:
Yeah, so I'll provide, again, my two cents in terms of where I think there's value, but also, I think, where there's probably still going to be limitations. And of course no one knows for certain, it's weather, but these are my thoughts.
So certainly, first of all, I think there's a lot of value-add in terms of these new AI forecasts and their ability to help us make better forecasts over the next several hours and into several days into the future. And I think anytime a meteorologist, and certainly the end users from our products have a new forecast model that's somewhat independent than the existing models and that has similar or maybe even slightly better skill, that is good for everyone. That is good for the meteorologist. Again, it's good for the folks who are at the other end of this because ultimately, we should have better skill and be able to give more lead time on those fire risk in Southern California, the hurricanes on the Gulf Coast. So again, it's a win-win situation. And I suspect, just like with the numerical models, we'll continue to see some improvement over the coming years.
Now, if the question is, and sometimes maybe this gets lost in the messaging, two weeks from now, two months from now, a year from now, are we going to be making those like the forecast we talked about where on July 15th, we now have hourly forecast of temperature and precipitation for a point location? I don't think we're going to be there, and I don't think still we're going to be there for a long time.
The reasons why are some of the things we talked about at the beginning of this podcast, which is these models, these AI models are still based off what humans devise, and it's really from the initial conditions of these numerical models, and we talked about the initial conditions of those numerical models are still flawed. We still don't have observations, complete global coverage. We probably won't for a long, long, long time. And even if we did have perfect global coverage, even if every point on earth we had coverage, we would still need precise information and accurate information.
And again, is that going to happen in the near future? Probably not. And so the chaotic nature of weather means that once you go out days, weeks, months, I think you're still going to be limited. So incremental improvement is what I would expect, and I think we're going to do better because of these AI models, but I don't think we're going to see this profound, all of a sudden, hey, now we can forecast out five months the same way we were doing one month out today.
Lance Glinn:
So Dave, weather is really ever-changing. And we talked about the one to 10-day forecast before, and it's crazy. You could have a 10-day forecast, they could put something out for day 10, and by the time you get to day 10, it could be the exact opposite of what was originally forecasted.
So how do you deal with the uncertainty of meteorology as it's obviously something so dynamic and something that really is at times very unpredictable?
Dave Margolin:
Yeah, so a couple of things have worked well for me and are things that I've had to learn the hard way too over the years. So one, I think, is that you have to have, or as meteorologists, forecasters, it's beneficial to have a repeatable process. And I say this because again, this is where human psychology gets involved, recency bias, things like that, stubbornness. It can take us off the rails. But the more you can have a process where, I can't remember the book where I read this, but it's forecast, measure, revise, repeat, and just have that cycle, that methodology of doing it, I think we're going to be better off.
The other factor here, and I think this is where our entire community can do a better job, is transparency in terms of our forecast methods, but also in particular, in terms of our forecast performance in a way that people will understand. Going back to the 10-day forecast you mentioned earlier, how often do you see when you're seeing that 10-day forecast, whether it's coming from your phone, internet, TV, the meteorologists or the service provide the accuracy, the likely accuracy of that 10-day forecast? Well, my guess is that that 10-day forecast probably has just very low skill, depending on the location and time of year, but relatively low skill above, say, throwing a dart or just using climatology, but they don't tell you that.
And I think it's disservice, it can be misleading, it can lead to problems, again, with the decision-maker, whether you're the decision-maker who's running a company and trying to determine whether they have to take mitigating risks to hurricane or whether it's a homeowner. And so I think we need to do a better job, we should do a better job at providing, again, those risk forecasts, those probability forecasts, and transparency in terms of the skill that we have for these forecasts that we're publishing, especially these forecasts that sometimes go out three to six months into the future.
Lance Glinn:
So Dave, as we wrap up our conversation, just how do you see meteorology evolving in the coming years and what do you think this evolution or how do you think this evolution could ultimately impact the financial markets and what you do at ICE moving forward?
Dave Margolin:
Well, I mean, I think it's an exciting time, and I know I'm biased to be a meteorologist. And I think for the next generation of atmospheric scientists and meteorologists, some of the things we talk about, there's just been such an explosion of weather services, data, great new models that are available, and it's making forecasts, it may sometimes not seem that way, but the forecasts are getting better.
And I think I'm excited about our ability not just to have the better forecast over the next 5, 10, 20 years, but what we talked about, to work closely, more closely and better with those folks like the emergency managers, the people who are on the front lines. And I think that's an exciting opportunity to do things a lot better and I think that is a direction I hope and think that we'll be going down here.
Lance Glinn:
Well, Dave, it was a pleasure having you join us on the podcast. I'll, of course, still continue to check my weather every morning when I get up. Thanks so much for joining us Inside the ICE House.
Dave Margolin:
Thank you, Lance.
Speaker 1:
That's our conversation for this week. Remember to rate, review, and subscribe wherever you listen, and follow us on X @ICEHousePodcast. From the New York Stock Exchange, we'll talk to you again next week Inside the ICE House.
Information contained in this podcast was obtained in part from publicly available sources and not independently verified. Neither ICE nor its affiliates make any representations or warranties, expressed or implied, as to the accuracy or completeness of the information, and do not sponsor, approve, or endorse any of the content herein, all of which is presented solely for informational and educational purposes. Nothing herein constitutes an offer to sell, a solicitation of an offer to buy any security, or a recommendation of any security or trading practice. Some portions of the preceding conversation may have been edited for the purpose of length or clarity.
