Now we'll move to our non-insurance businesses. I'll start with BNSF. As I've highlighted, a number of our non-insurance businesses provide critical product services; BNSF is a great example of that. 32,500 miles of track in the West, moving core products for a number of customers that touch every industry in the US. You'll see the results... some improvement there. But what we really want to highlight today, and Katie will be joining me on stage and she'll touch on this, is that we have a lot of work we know to do at BNSF. We have a great group of employees who have been working very hard—I would say boots on the ground. So, you've heard me talk in the past that we have to work hard in our yards and work on how we can move our cars quicker and meet our customers' expectations. And we're doing a very good job on the customer service side, but we've recognized we've got to get better operationally. Our team's also been very focused on what resources do we have—do we have too many locomotives? Because actually too many locomotives, it sounds counterintuitive, but can be a problem. One, you're not as efficient, but the congestion and everything that comes with it... so our team's been very focused on that. And then how do we best use our employees? Well, that's something our team's working hard on to become more efficient and more effective. But we also have to very much recognize where we are versus our industry peers. This is the six Class I railroads that operate in the US, and we're one of them. And you can see that last year we were fifth out of six, and that's a reality of where we are. But we're also getting better, and we are going to get better. We recognize that this performance... our teams have worked hard, but there's a lot of room for improvement.
Now the good news is if I look at it in 2025, and what we have here is our operating margin. So the 34.5% you see for Berkshire, that's the operating profit that came back to BNSF associated with its underlying operations. That operating margin improved by 250 basis points. That's a very positive outcome, obviously, and by the way, in fairness to our team, that's the work they've been putting in—that was, on a nominal basis, the largest improvement across our five peers. So we're pleased with that, but we know there's a lot of work to be done. If you look at our first quarter results, happy to report that, okay, we went from fifth to fourth, but our team would be the first to say there's a lot more to be done. And if you look at our overall operating margin there, it's very consistent with the result last year, and the efficiency we've delivered is being maintained and improved over the quarter of 2025. So again, we see a lot of opportunity here to continue to get better, but to achieve where, say, Union Pacific is as a leader with an operating margin of 39.5%, we know that's going to require a step-change both in how we're operating and even how we approach our operations. An important step that we've identified—we like to identify the gaps and where we can get better—is technology, and we're doing a lot at BNSF, and I'll touch on our other businesses here when I go through technology. That's where we see a step-change, or potentially where a few of our peers have gapped out versus how we're using technology.
I'm going to back up to GEICO and then I'll come back to BNSF, because it was approximately four years ago I was in a GEICO meeting with our management team there, and they were discussing this price to risk and segmenting customers. We had our operational team from GEICO; they had the commercial team, but they had the tech team there, and you're always looking for some help. But what I heard in that discussion was a clear technology transformation that was happening at GEICO. It was obvious that technology was going to be a big part of the solution as GEICO tackled their certain challenges. And as that meeting wrapped up, I very much wanted to spend more time with the technology team to understand what was driving this—and they were calling it a technology transformation—because I could see that it was so applicable to what we needed to do and what we would pursue across our non-insurance businesses.
So what is this technology transformation that they described at GEICO? I'll summarize it in a few different ways, but first and foremost, we recognized we were going to become a builder of technology rather than just a buyer of technology. And that meant that instead of having a number of systems and often buying the related applications or software that came with it... and yes, it's a valued application, but it was disconnected from all our systems. Obviously, we didn't have that ability to then use the information, get to the data. And what they started to talk about is simplifying the infrastructure, making sure we would build what we needed ourselves and deliver solutions back to our customers, and we would have clear access to the data. All things that make a lot of sense, but it's a massive challenge and it doesn't happen overnight. And we're still on that journey at GEICO in year five.
There's no question, but I quickly recognized that this could be used across our other businesses. Very fortunate that at GEICO they had put their leadership team in place to drive forward this transformation, and the most senior leader then came from GEICO, joined our non-insurance operations, took on a senior leadership role—the leadership role helping us with the technology transformation at Berkshire Hathaway Energy and then also as a Senior Technology Officer at BNSF. So we started down that journey, and one of the first things you have to do is say, "Okay, we need a different resource base". So now we're hiring engineers, we hire developers in our technology group that help us start to build the solutions we need for these businesses, and it's going beyond GEICO now. And we still have our valued employees there, and they may be retraining or transitioning to other roles, but the reality is we need fewer people managing the applications and the software and more people building outcomes that our businesses need.
Now, when I asked our team, "Well, how does AI fit into this—Artificial Intelligence"? Well, it's actually a big piece of this because it's effectively what goes on top of a lot of our systems, and that's what they're building. They're using AI to build applications, and that's all great, but we also know there's certain risks around humanity. There's risks broader globally and for our country, but there's also risk within our businesses. And as I just start heading down this path, I said, "Well, okay, how should we think about this? How should we all be comfortable we're approaching this correctly"? And they said, "Well, we don't really like to call it Artificial Intelligence; they call it Narrow Artificial Intelligence," and they have three really important principles associated with it. The first one was that yes, we're using it and we'll use it with these engineers we have and these highly skilled individuals we brought in, but how are you going to manage it? Well, the first thing was that we still have our employees—our senior management team—involved in implementing the recommendations that we then receive associated with the architecture or the framework they put in place. There may be things that still occur and should occur just like they did within our systems, but as it moves up and the important decisions are being made, there's human involvement; our managers, our employees are involved and that's part of the governance that's effectively in place.
Greg Abel: The second piece is what they call the safeguard, and the safeguard is very intriguing because right away, of course, we all want good governance. We want that in place, but what's that mean? And our team said, "Okay, here's how I would describe it: if we ask for an outcome—we want a recommendation or an action—and we ask it now, and then half an hour later we ask, do we get the exact same outcome?" If we can receive that same outcome, it's effectively the safeguard. We know we're utilizing that application properly. And importantly, it means we've got a defined data set that we're comfortable with. I like to call it the constraint. We know we're constraining our data; we know what data we're using and we know what data is coming in. Now, when you talk about all the operations that we're focused on, yes, the next day of operations comes in and it updates that data set, and we may get a different result if we ask the question the day later—it's got new information. But if we ask it, "Well, if you ignore today's information and just focused on yesterday, do we get the same answer?" Yes. So we call that our safeguard.
And then the third thing on technology and associated with this Narrow AI is it has to be additive to our businesses. We're not going to do AI for the sake of AI. You can spend a lot of money in this area, and we need to know what we're trying to achieve and do we see a valued proposition for the businesses. So that's what we call Narrow AI.
And if you see how it's starting to be applied at BNSF, it's incredible. If I think of BNSF, we have the expansive network I touched on. We have a variety of trains leaving from a variety of points every day. I've touched on it—it can be the intermodal trains of 150 to 200 on our tracks a day, which are moving very quickly and often leaving LA to deliver product in Chicago 48 hours later. Or it can be—in the last quarter we had more than 750 trains a day moving across that system. There's weather, or there's equipment failures; we share our tracks, we allow Amtrak to use them, they can be running on time or they can be running behind schedule. We have to adjust all to that. And the reality is Katie and her team have a system that's been running for 177 years, but we were not there in how we could use technology to operate that better. And that's what we're using, or we've just started down the path of—that's how we know we'll see that step-change in our operating performance.
Now to summarize all that, but I want to let you know it's all around operational excellence. We are going to get better at rail, but we're going to use that framework across all of our businesses. They very much will create the framework and then our teams can embrace it if they so choose, and we'll help them see the value of it. But there is an opportunity there, and I'll break it down with one last comment around technology. When you think of Artificial Intelligence, everybody talks about the Large Language Models, and okay, they're learning models and there's a lot more to it than I just highlighted. But I summarize it as one thing, and this is why there's an opportunity across all our businesses: those Large Language Models, I really communicate them—and I communicate them to our teams, or at least it helps me understand it—they're "Large Logic Models". We're at this point in time using it to solve logical challenges in our business, and what we're trying to do it is in a more efficient fashion—i.e., do it more quickly and get to a better answer. So that was a lot in technology, but it touches the whole franchise of Berkshire.
If I move to energy now and provide an update there, I'm just going to touch on the opportunity first, then come to the challenges. Because as I've just discussed technology, that's the opportunity in energy: one of the core inputs to all those data centers and hyperscalers associated with Artificial Intelligence is energy. Our businesses have that opportunity in front of them at Berkshire Hathaway Energy. And yes, we're pursuing them, and we'll do it—I'll touch on it in a way we view as the right approach for both our states and our customers. But I would highlight it's not new to us. If you just go across the river a little bit east to Iowa, we serve just under 50% of that state. If you look at the number of data centers and hyperscalers in that state, it's very significant. We have four very large hyperscalers, data centers there, or builders of them, and ultimately their customers using it. But if I look at our peak load—i.e., the amount of energy being used from those data centers—it's at 8% of our peak load.
And the only reason I highlight that 8% is when I hear people in the industry and all the utilities around us, a lot of states, they're talking about this great opportunity and, "Geez, hopefully in the next five years they'll be from a relatively starting point—they want to get to the 5 to 10%". And we're already at 8% and we see opportunities to grow that by 50% over the next five years, or potentially more. But we'll do it in a way—and you're starting to hear more and more of this across the US—we'll do it in a way where we're not going to impact the costs of our other customers. These users—the hyperscalers, the data centers, and the users of the energy—they have to bear their full cost. We can't transfer that burden across all our other customers. And that's a principle we've applied across all our utilities and from the very early goings when we're building these data centers, and I would highlight I think our team's doing an exceptional job of that.
If I again go back to MidAmerican, if you look at their... with all the data centers and hyperscalers and the infrastructure they put in place, their rates are still 45% below the national average. That's just unheard of. It's an exceptional outcome and it just highlights they're doing the right things when they build this infrastructure, or it's a part of it, and we would highlight we have similar positive outcomes across the rest of our utilities. Now, and I would note one other thing: our gas network or our infrastructure there, our large pipeline company we have there—as they build out all this infrastructure not just in our utilities but across the US, our pipeline footprint will growth. A lot of it's being built by natural gas, and we'll meet that challenge. But here's an interesting point: 15% of the natural gas consumed in the United States is touched by our pipeline network or one of our core assets there. So again, that's the opportunity on the energy side.
But it's not without its challenges, and we've talked about this the past few years. When I think of the Berkshire Hathaway Energy Group, what's the challenge? It's what I call the regulatory compact. We leave your capital, our owners' capital, Berkshire's capital, in these businesses and often a portion of the earnings that they generate, we may reinvest back into those businesses, and for that we get a very specific set of return. And it's a fair—over the long run it's been a very balanced and fair return—but how do you measure that? It's versus the risks we take on in that business, and that's the compact: "Okay, you're going to pay us X percent return, and what risks are you asking us to take?" And that model has worked very good for a number of years and for centuries, but the problem is it's becoming more stressed if you think of inflation.
If you think of the data center challenges—but I strongly believe we're managing that separately—and then you move to assets that are 60 to 100 years old that are starting to retire and we bring those into the network. The challenge is every day to get more efficient, more effective from the operational side. But as a regulator, as a governor, you're very focused on, "I don't want my rates to go up; I don't want to take on more risk". They want to transfer that back to us, and that's the regulatory compact. And unless that exists—if we don't see that balance—we do not, if we don't see that balance, we don't deploy our capital back into those businesses or into those utilities, and we work hard to maintain it.
But there's been a very important challenge we've had within that we've touched on in the past too, and that's wildfires. Wildfires in the West, very prevalent the last 15-plus years in California. We experienced a very significant wildfire in Oregon in 2020—or a number of wildfires across Oregon—but we, being the state but also the company. So there were a number of wildfires across the state. We had certain equipment, certain high winds, we had certain failures with our equipment that contributed to those fires. And associated with that, we fully acknowledged where there was causation and where we were responsible for it. But there was also associated with some of the fires, and specifically one fire, a class-action lawsuit that was um had very large claims against our utility there, PacifiCorp.
And we had to approach it such that we'd resolve all the other matters, but that was a class action. And there was specifically one fire that we strongly felt we weren't responsible for—there was zero causation. There was an Oregon Forestry Department report that said though that PacifiCorp did not contribute nor cause the fire. We took a very strong position there that one, we were not going to put more capital in to fund the entity in these type of risks and these type of obligations, and secondly, we would challenge that liability verdict. And we challenged it; it's been a long process, but as owners and shareholders—and this was a very significant event that occurred in this past quarter or occurred in April—we're very fortunate that it was up to the appellate court. They reversed and remanded that liability verdict and said back to ground zero; start over again.
And what they were really saying was that that class of customers and who did we actually affect and where was the causation... that's been that will be revisited and then the related damages. Some positive things associated with it: we recover a billion dollars of security we've already posted. The people, the law firms that pursued it, are responsible for our costs associated with that period of time—not our litigation costs, but the costs we incurred in posting the bonds or posting that security—that's $10 million or approaching likely $10 million. So, but the most important thing is we've reset the stage there and that's very important because we're working hard to get that regulatory compact balanced and getting the right outcome, and we do want to see these utilities move forward and we want to be a very good operator and steward of those assets for our customers.
So, the last thing I'll just touch on regarding wildfires: so when you think of PacifiCorp, yes, we've addressed that challenge, but to get the right compact, we've worked with Wyoming, Idaho, I've touched on Utah on this stage to say it requires a judicial system that supports the legislature, the laws in place, but more importantly, we all, or as importantly, we need good legislation that then sets that balance. We've had it across those states and we'll continue to work hard across our other states. So, an exceptional outcome and wanted to make sure... you know, there's still a lot to be done there because we're back to the very first, back to first base on the legal proceedings.
Now, moving to our manufacturing and servicing businesses. This highlights our manufacturing group, that represents approximately 70% of that group, and the service and retailing groups in the gray or beige. I like to think of—when you think of our manufacturing group, we've got three groups there. We have our Industrial group, we have our Building Products group, and we have our Consumer Products. The consumer products, servicing, and retailing, as I've touched on, is now under Adam Johnson. We're fortunate to have Adam as our leader there; he's managing 32 of those companies and we'll have him on stage and we'll expand on that more. If I go back to those a few of those core manufacturing groups, I'll start with the Industrial group. And even when I think of the Industrial group, I like to break it down into a couple other groups, but it's a good way to think of our businesses, and that's why I want to share it. Within the Industrial group, we have a Metals group that is very strong; there's three businesses.
We have Precision Castparts; it's a business we acquired 10 years ago in 2016. It's run by Mark Donegan who was the CEO when we acquired the business and he continues to run it today. And as owners and shareholders, we're very fortunate to have Mark in that position. He understands Precision Castparts inside and out, he understands the industry, and very much works towards delivering solutions for our customers.
The second important part of that Metals group is a business called IMC, and they are an international metalworking company, and it's really interesting to see that company—one, they make the tools that remove steel. So they'll take a cylinder of steel, they create the tools, and then that gets utilized in a variety of other industries. It can be the aerospace industry, like a Precision Castparts and I'll touch on that, or it can be another industry like the auto. If you think of what's happening in the aerospace industry—and this is why Precision Castparts and IMC have such a significant backlog, or I'm highlighting a backlog.
If we look at what Boeing just announced last quarter—or this quarter, but just recently—their number of planes that they delivered went up by 11% quarter-on-quarter; that's phenomenal, and they're talking about even doing more. Very similar results at Airbus, and that's who Precision Castparts serves and also often IMC serves that industry. And that's remarkable. But if you hear of the backlog in this space, it's 10 years. And I did ask our team the simple question, I go, "Well, is that many more people really flying?" Like, I get it, we're post-COVID and it's building up. And I sort of obviously knew part of the answer, but it's really remarkable why there is that demand. And a lot of us know this, but the reality is to see what's driving it is the efficiency of those planes and engines is so great now that it's better to buy the new plane and retire the old plane.
And what you have is this 10-year backlog that we're seeing, a very similar backlog across our metal businesses when you touch on a Precision Castparts or IMC. Now the third piece of the Metal groups—and by the way I should just touch on this—we acquired IMC basically 10 years before Precision Castparts, so you go back to a 2006 timeline, we acquired 80% of it. Again, we're very fortunate to have the senior leader there, Jacob Harpaz, who was at the business, the senior leader running it back then and still runs it today. If I look at how Precision Castparts and IMC work together, Precision Castparts is now, if not the top, very likely IMC's number one customer. We have them working on joint solutions.
Now move to that third group: in 2022, we acquired Alleghany and we're fortunate to have that in the family now and a very good addition. But along with it came three non-insurance businesses; there were a variety of other ones, they're tucked into the appropriate place in other businesses, but one of them that stood alone was W&W|AFCO Steel. And it was a family-founded company; it had transitioned to Rick Cooper—he's here over in the manager section, he's the CEO. And it's a remarkable business: they create basically, they contribute steel into a variety of core infrastructures. It can be bridges, it can be stadiums, it can be arenas. Their most famous one is the Las Vegas Sphere. And here we bought an insurance company—and Warren has touched on this—we sort of had these nice add-ons that I'm not sure we spent a lot of time valuing that side of it, but incredible additions to the Berkshire family, and that really comprises the Metals group.
Now, if we move to the second group, it is the Chemical group we have within the industrial sector. We have three of those: we have Lubrizol, going back to 2011—we've touched on that business many times. We then acquired OxyChem last year; associated with OxyChem, we announced the acquisition and we closed it on January 2nd. Very nice addition; I would say they provide, they produce two core commodities, so it's more a commodity chemical business, and they're valued commodities in the industrial sector, but their plants can't be replicated. That would not be easy. So we've got valuable assets.
And then the third piece of the Chemical group is a company called LSPI. And I'm just highlighting because it's a real gem for our owner-shareholders: what that does is it creates a drag reduction agent that allows oil to move through a pipeline. And you can imagine in the environment we're in right now with the fundamental supply and demand imbalance on oil, the more oil that can be moved through those pipelines—and you can't quite double it, but you can get darn close. They have an amazing product and obviously in very high demand.
Now, the last thing I'll just touch on with the Industrial group is we have Marmon—excellent business. And the reason I'm touching on it at the end: it's really amazing because it's the catchall. It touches our rail industry, it touches the energy industry, it touches the metal industry, it touches the chemical industry. It touches in all the core industries in the US, and again a remarkable asset that we have and will continue to create strong value for our owners and shareholders.
Now, the second piece of the manufacturing group is our Building Products group. I'm not going to go through all the businesses in there because we have one that is the bellwether and it tells you how the rest of them are doing: that's Clayton Homes. The other businesses' results follow very closely to that because with Clayton Homes we're building manufactured homes or site-built homes, and there's a lot that goes into it, and our other companies provide both products to them and across that space. So it could be the insulation, paint, carpet, a variety of other things. If you look at Clayton's results, if you go to the manufactured side of the business, our results are down on true homes manufactured and sold—down approximately 10%. A little better than the industry average, but that gives you a feel for it. And if you go to the site-built, i.e., where we were a home builder, they're down around 5%. And the numbers I've been seeing for the last quarter are probably more like 7% across the industry. And that's obviously driven by where interest rates are and certain other challenges for the consumer.
