Honeywell International Inc. (HON) Earnings Call Transcript & Summary

Again, we're very excited to have Honeywell with us today. We've got Vimal Kapur, who is the Chairman and CEO of Honeywell. And Mike Stepniak in the audience, he's the SVP and CFO. And Vimal I walk over, it's only been a couple of weeks since you reported earnings. We know that. But maybe give us an update. You do have a fair amount of shorter-cycle businesses. Obviously, Building Automation has been very strong. Aero has been very strong in your long-cycle businesses, and you've got a reasonably strong orders.

So take us through what's been happening in Q1, if anything, is different, new, anything like that?

I think fundamentally, the momentum what we saw in '25 continues in '26. If you ask me what's the change year-over-year, I think external markets remain very, very similar. Aero, very strong. Building Automation, we have a strong, both short and long cycle. Industrial Automation business actually is doing quite well in North America, but not that strong demand in Europe and China, which makes our guide more around what we guided there. Essentially, the only market where we see more lack of demand is petrochemicals catalyst. And I think the world has enough capacity. We're talking about it probably for the last 2 or 3 earnings calls. And our guide assumes that the situation persists in 2026. But the opposite tail of that is the long cycle and process is very strong. Our orders have grown 2 quarters in a row. We even see Q1 shaping up quite well. So long cycle capacity build continues on LNG and refining, while short cycle catalyst demand remains, specifically on the petrochemical side, more dormant to flattish because of excess capacity. So we feel good about the guide we provided about 3 weeks back for the year and including for Q1, things are shaping as directed, and we'll work very hard to be on the upper end of our guidance.

Vimal, let me follow up on one of the things you just said, like just in terms of projects on the long cycle side. Like do you see customers in this environment actually like going forward with the projects. You're getting the orders, right? You've talked about projects actually going to revenue as you get later in the year. Do you see that happening more regularly? Or...

I think what has happened is, given the -- if you just see the cycle of industrial over the last 3 years, there are some segments which are heavy, very high demand. Aerospace, data centers, LNG. So what I have seen is that because of the high demand, customers are now also placing orders more ahead of the time compared -- due to capacity constraint. Let's take example of our LNG business. We are booked till practically end of '27 already and actually early '28. So order booking keeps growing, but our capacity is limited to a certain volume. And aerospace is a similar example we've been talking about. Aerospace is booked for a long time, but we can only grow our volume by 12% to 14% every quarter. So I think these bookings are reflective of the long demand in some of these sectors, but the volume growth or revenue growth is constrained by capacity, which exists to actually physically deliver the volume what we have in our backlog. And we're raising our capacity across the board, where aerospace is investing more capital to increase its capacity. We are expanding our facility for LNG facility by 2x because that's the level of demand. But in spite of that, we are booked. So that kind of shows the demand profile there at this point.

We'll definitely dig in a little bit more on that, Vimal, later. But I want to ask you like, obviously, there's a big few months for you as you get closer to sort of separation. I think you tell me, but one of the main reasons to separate Honeywell was to improve focus across major businesses. And so I guess the way I want to ask you this is, have you seen any of those positive impacts already? You've already separated Advanced Materials, became Solstice. Maybe one of the biggest beneficiary if you look at a small or streamlined Honeywell in terms of the businesses?

I think across the board, I mean, as I spend more time on RemainCo, our ability to look at each segment and its growth potential is very different than what it was when I started as a CEO about 2.5, 3 years back. And I think fundamentally, if you see how decisions we made on Industrial Automation since we decided the spin to say these segments fit less. By separating them, we become a pure-play sensing and measurement in industrial is a new development we did over the last 6 to 9 months. So every segment, we are able to bring strategic clarity, but also how as one company, we are going to work as a much more effective automation player. So there's a -- if you ask me when we made a decision to separate as 3 companies, which was October of '24, when we first announced Advanced Materials as Solstice. And today, 18 months, actually, my confidence is very higher because now we are living that life. In October '24, it was a hypothesis. We think it's going to be great. The charts look great. And our research showed this will make a lot of sense. Now we are living that life. Aero is living that life. They see more potential, more opportunities. Automation sees far more opportunities. We are also, timing-wise, spinning as an automation company at a time when AI is becoming much more a tailwind for us in terms of creating new offerings, and we'll talk about it. That's just a nice timing-wise, the convergence is happening at the right time. So fundamentally, I'm a firm believer, focus happens. Focus definitely benefits us. Simpler companies are easier to run and easier to execute, and that's the reason we did the split here.

The AI common is a good advertisement for everyone. So we'll stay tuned for that. But maybe I could ask you about new product intros because '25 was a good year for that. You mentioned 4% growth. But I think you've talked about it was weighted toward Building Automation. So maybe give color as to why NPI has been able to build out there faster and how it then would spread to the rest of the organization.

Look, our effort to make new product as a primary vehicle for growth is across the board, Honeywell. This is not a specific strategy limited to one segment. But we have seen more results there. We will see more results in Industrial Automation starting from second half of the year. Just cycle. It's the cycle of developing new product, specifically in case of Industrial Automation, a lot of our products are certified by customer. So I developing alone is just half the battle. And then somebody has to go through the cycle of approving. It's just -- Buildings have a lesser hurdle rate. Industrial has a slightly longer hurdle rate. Now in the process market, the situation is a bit interesting, and we are not seeing the outcome of new products, primarily driven by the end market dynamics, less by our own innovation. So we invested a lot of dollars to create new offerings for biofuels, for example. And now the biofuel adoption has gone practically 0 for the last 12 to 18 months. So did we invest in the wrong place? I absolutely don't think so. I think the world needs it. But now adoption has gone kicked to the right. And the solution businesses, the NPI cycle is more cyclical because of nature of the solutions because it's a projects business. BA, Building Automation and Industrial Automation are pure-play product businesses. It's pretty black and white. If they grow, Building Automation grew 7% last year, 4% was product, new products, 3% was price. And we think we can repeat the same logic much more easily in Industrial Automation. And the process is a solution business. It will always have an end market driver. If end markets are favorable, we'll grow there also at the same rate. So strategy is across the board. Our R&D is now above median. We spend a lot of offering on offering management, customer co-creation. So that's a central thesis of how Honeywell wants to live its future.

Got it. That's very helpful. And Vimal, I remember when you and I first talked, you're CEO for not that long, you talked about sort of Honeywell trying to improve data collection and information gathering because that would help you price. And so I thought it was really encouraging to see 4% price in '25. I think you're guiding to something similar in '26. So maybe talk about where you are in terms of maturation towards a full dynamic pricing? And what could that mean towards your incremental margin moving forward?

I mean, look, the pricing has become -- the inflation has become a, I would say, a new gating factor for industrial company like us. And inflation is persistent in that rate of 3% to 3.5%. And categories keep changing, but output doesn't change. It just remains in that range. So labor cost keeps increasing, which is a big category we buy. Electronics is more expensive. Capacity of semis. We all know about memory. We don't have much impact on that, but still overall indexation perspective, the cost is increasing. And commodities prices increasing. All commodities are expensive: copper, zinc, gold, platinum. We use a lot of precious metal for making catalysts. So the input costs increasing, the first and foremost thing we are doing differently for the last 1 year is talk to customers about inflation in every meeting. I've been working for nearly 40 years. I did not grow up to talk about inflation. Our customers, you talk about new products and delivery and how things are going. But now we talk about inflation because we are conditioning market on what's coming ahead. So when we raise prices, they are not really surprised. So I think one thing we are doing differently is much higher communication. Second is we are highly sensitive on elasticity and price volume. Last year, our price was 3%, volume was 3%. This year, guide is again 3% to 6% upper end. It means basically the same. If we have to get volume, we have to be sensitive that we can't throw price at people and assume they're going to absorb it all. And memories are long. And we just don't want to be seen as a irresponsible player who's just passing on all our problems with the customer. So productivity has to work equally so that I have some optionality not to pass on all prices. So price cost, if it's not neutral, if it is 10 or 20 basis point negative. But if I'm getting volume and volume leverage, it's still a good news because our margins are expanding. So I think we are putting far more effort on productivity execution apart from pricing. And finally, I would say the thing which gives me most confidence to deliver pricing is through new products. Because if inflation is here to persist, if it's going to persist in '27 and '28, there's only so many years we can keep going and ask for 3%, 4% price increase. We are anticipating that will have a demand effect. However, if we can create a higher value capture for our customers through our channel partners, through our OE, then price becomes less of a debate, then the value becomes in the front of the discussion. So overarching, I would say our playbook is very different compared to 2 years back, much higher level of customer communication, much higher level of sophistication on elasticity, much higher focus on new products. And all in all, we feel confident. I mean this year, pricing will be again 3% range, 3% to 4% range. We'll see how the year progresses. And I think it's going to be same in '27, too. I don't see -- I haven't read any news by which I believe that industrial inflation is going to go down based upon the commodities and the products and services we buy.

Right. And Vimal, you mentioned limited exposure to memory pricing. So like you're fine with that and you can stay ahead generally on price versus cost?

Yes, it's limited to just one business. We have the business of productivity solution. We make mobile computers and scanners. Those are like iPhone that has onboard memory. And that memory is something we have to work with the memory providers, which is Samsung and Micron. The good news is they are both our customers because they buy gas detection products from us. It just makes it easier. We know we have a deep relationship with them for many years. So it has been helpful to understand what they want us to do to become a more trusted supplier, and they made it quite transparent to us. They essentially want us to migrate to more standard products. The volume which they are producing, they'll say, if you start buying this, you won't have any issue, but we have to redesign our product to migrate towards those memory, which is more of a transition thing. So I feel good about protecting our volumes this year. And longer term, we will manage it effectively. But the issue is narrowly into one business, not widely spread at this point.

Helpful, Vimal. So another initiative I think you spent a lot of time on is what I think you've called compounding business models. So you take the best-in-class business in Honeywell, you turn their ability to harvest their installed base and you kind of translate that to your other businesses. So sort of where are you in that process? Because again, it seems like it's worked very well in Building Automation. You tell me, and maybe it's still on the comm for other businesses.

Fundamentally, I think one of the things we had to take a call is as we become a pure-play automation company in 6 months or less, how do we define what automation segment we play? Because automation is a $500 billion, $600 billion TAM, and we're going to be just $20 billion. So the line we have drawn is 2 ways. One, we want to be in the mission-critical automation segments where product or solution matters with the customer because that gives you a longevity of relationship, that gives you pricing power, that gives you aftermarket because the stuff matters to them. The second, we want to live with the business model of build and mine installed base. Now once we define those principles, it's easier than to say what you want to own and what you don't want to own because that does fit in this definition. Now on a broadly speaking, at the exit, our total services and software aftermarket is about 40%. And having lived with aerospace, they're already at 50%. So a question we ask is, how many years will it take to get this mix of half and half. And I think it just derisks the business model. I'd say the maturity is medium across the board in Honeywell because we never emphasized the business model being the primary driver of our existence. We have services business already. So that's not a novelty. All our competition has a services business. So there is no ingenuity in that. But when you say my business model is aftermarket services, you do things differently. I'll give you 2 examples. We put in a system in place starting, I would say, late of 2024 and entire '25 to record all our installed base in a single system of record in entire Honeywell. So I know all the installed base of all the businesses by number of assets, by customer name anywhere in the planet. Now why it matters? Because we already know our penetration rates. So we can ask a business to say, why your penetration is low? What are you doing? I mean, why it's only 30%? Your colleagues have 60%. So you guys don't know what to sell, you don't have offerings. So we did not think of those kind of systems. Similar example, it will be all our effort on our Forge platform is essentially to mine installed base at the heart of it. We want to mine installed base in a structured manner using data science, versus a good old way of break, fix. Okay, something fixed, give me a call, I have a service contract. Yes, that's still in play. But the world is changing fast and people expectations are changing. So the business model focus is making us behave differently. I can see people coming back to me with the new ideas and how can they generate new aftermarket service, which was not talked earlier. And I think that notion will play over a period of time. And we'll talk during Investor Day, when do we think we get from 40% to 45%. I mean it's a time-bound thing. It's not going to be immediately, but we do believe that mix change is a big favorable factor for us moving forward.

That's very interesting. And so I want to shift gears and talk about Quantinuum for a second. I think you gave a lot of good color on the earnings call, but it definitely seems like the momentum in that business picked up significantly. So was it that you reached an inflection point in the technology last year? How would you characterize the acceleration activity at Quantinuum?

I think the 2 things are helping us there. First of all, the hardware platform, we launched a hardware platform in November of 2025, which generates a 48 logical qubits at a 99% plus fidelity. So I know lot of jargons here, and that is intense. So let me demystify that, what exactly it means. When a quantum computer has to be taught in terms of its accuracy of its logical qubit. So physical qubits don't really matter if your error rate is 50%, it means everything you produce is wrong. So we have solved the fidelity. It means what we solve for is very accurate. So our -- for every 2 physical qubit, we can produce 1 logical qubit, that's a big deal in quantum. Now in about a year's time from now, we will have a quantum computer with 100 logical qubits. That's happening with high certainty. At that point, it's more powerful than any other classical computer available in the world. That fact is now making customer more intrigued to what to do with it because it's coming in 12 months. It's not future. It's not somewhere out. And if I'm in a business where my business can get impacted, I need to do something about it. I can't be sitting on the sidelines. And the 2 most impacted -- 2 most interested end markets are banks and pharmaceutical. So molecule discovery can happen faster. They can do multiple state discovery simultaneously versus experimentation. But certainly, those use cases have to be proven. So we'll be working with -- we are working with different pharmaceutical companies to develop use cases. Banks are looking at a different way to look at their encryption, which is a big deal for them because quantum provides a very higher degree of encryption capability, which doesn't exist in a normal way. So the short version of the story is as we see traction of customers engaging with us and actually contracting with us even at a modest way, that gives us the confidence that this company is capable to be a stand-alone company, and Honeywell control is no more necessary. So the hardware platform has been accomplished. Business model has to be accomplished in next, I would say, 12 to 36 months and that's our window, earlier the better, because we don't want to keep ownership of that beyond a point of time. The other thing which is helping us, apart from the hardware proving, I mentioned 2 factors that are helping us. Second is a word called AI because AI needs high compute. And one of the things we have been working with NVIDIA since last year is how to create a software environment in which workload is shared between GPU and quantum because if customers are doing an AI-based use case development and if compute power is a constraint, then we want to share the workload here. So that also generates customers' interest because they're already doing it. I mean we're not competing with anybody there. We kind of -- it's a coexistence. Like GPU exists along with CPU. It's not that when GPU came, CPU doesn't disappear. So when quantum come, it doesn't mean other things don't exist. It's a coexistent strategy, which we believe is going to happen there. So I remain very optimistic. I think there's a lot of value capture for our share owners because Honeywell owns 52% of quantum. So at the exit, when we do it, it has a benefit of P&L pass-through, which is now $250 million of investment we make every year. So certainly, that gives us a relief. And on top of it, the cash it generates is certainly an upside for us, and we can create value for our shareowners.

It's very interesting, Vimal. So I want to open it up to the audience in a minute, but let me ask you, the other thing you did when you became CEO is you restarted the flywheel at Honeywell, as you know. And so one of the first big acquisitions you made was the Carrier Global Access Solutions business. So maybe it seems like it's doing really well inside Building Automation, but I'd be curious for you to update it and talk about what lessons you learned as this was your first big integration as CEO.

I mean I would say the fundamental lesson is that bolt-on acquisition is the best way for us to stay disciplined. We acquired in the spaces, which we know well. We already knew this asset for many, many years, and we were not dabbling into a new space. We exactly knew the space. We know the segment side. We know the customers. We know the cost synergies. We know the sales synergies. So our primary M&A strategy has been bolt-on. If you see the subsequent acquisitions we did like acquiring LNG business or Sundyne, these were bolt-on to UOP business, right? So required smaller tuck-ins of cybersecurity bolt-on to process automation business. So I think that has been our lesson learned that staying disciplined, both by strategy and by financial is the key. I think other learning which has been very encouraging for us is that Carrier acquisition ended up being a carve-out. And we had to become a sponsor to basically take just the business without anything else. So that gave us a new capability to act like a sponsor. So LNG business we acquired were also a carve-out from Air Products. So we've generally become comfortable to do carve-outs of our bigger assets. And that gives us more optionality. I mean, it's not that's a strategy, but you don't have to acquire the whole business. You go after what you think really is required with that. And final point I would say is that our ability to live with the business model for each of these acquisitions. They're all truly living to our spirit of mission criticality, build and mine installed base. So we are acquiring businesses which live to our -- what new Honeywell we want to build. And this is really helping us to reshape our portfolio. I mean, as a headline on exit, our net portfolio changes 30%. So 15% of the revenue has been exited and 15% of the revenue has been added. It's a pretty substantial shift. And we want to show that in Investor Day, one of the biggest goal I have is that people understand what's our portfolio now because I fully understand the change is significant. And probably everybody is tracking it one at a time, but we want to give a more holistic picture where we stand today.

Yes, that's interesting. Any questions from the audience? We have a question.

As AI data centers build scale, are you seeing architecture standardize across customers? Or are workloads driving more bespoke systems?

I mean the data center, our participation in data center is limited to our Building Automation business, right? And we think customers are talking much more of standardization of their build. And they want to make sure that they can copy and replicate that. So I would say that standardization is what we hear from all large builders. And the key thing they are making us do is to reduce the cycle time of execution. It turns out that we are the last guy or gal in into the commissioning. We are the last people to show up. So we are always behind because the project is behind. So we have been working with hyperscalers and understand that how we reduce commissioning time of our system by 90% using the tools of self-commissioning. So while they are standardizing, it's helping us to design our system so that our commissioning cycle can dramatically be reduced. So rather than we're taking 3 weeks, we can do it in 2 days. So that 2 weeks uptime is a lot. And we are actually launching that as we speak with one of the large data center developers. So standardization is what I hear all the time. They want us to have standard products. They want us to have standard execution. Honeywell's biggest value proposition actually for our customer is our global scale. Regardless of who you are, we exist everywhere. So it's a peace of mind. You can build a data center in Norway or in Indonesia or in Virginia, we can get it done for you with the same standard. If you have a standard, we are the company to go. If you do not have standard, you can shop around. So I think it works good for us from a value proposition perspective.

Any other questions?

Can I ask one more?

Yes.

On to quantum, from an industrial systems perspective, does quantum feel like a 3- to 5-year integration opportunity or more of a 10-year plus research arc?

I would argue it's more closer than far. I think we believe that quantum will have an impact into commercial world more in a 3-year window versus even in a 5-year window because the machine, which is capable of breaking the computational power will be available in approximately 12 months from now. What is not available is the capability to use the machine. It's a pretty strange situation that you have a road, but you don't have any car. I mean we don't know how to make use of it. So that time, the use case development or economic value creation is a limiting factor. Will it take, as I said, 12 months, 24 months, 36 months. But the interest level from customers is very, very high. And therefore, I -- my personal belief is it's more 3 years versus -- it's definitely not 10. And also, it's a good solution to this compute power need of the world because quantum doesn't need big space and a lot of power. You can fit 2 quantum machines in this room, just to give you a perspective. So it doesn't -- it's a very different technology. So it also is helpful to scale the compute power, which the world needs without endless need for expanding a lot of data centers.

So maybe, Vimal, just digging into the segments a little bit more. We've talked about Building Automation a lot, but I just have one follow-up there. Like your solutions business has been outperforming your products business for a while. And I think you've told us before that products is much higher margin. So does '26 get more in balance, like how do you feel about that?

It's definitely more or less. I think last year, we grew solution at 8% and product at 7%. The solution, our aftermarket services grew actually at a very high rate. That's because our connected building solution is becoming more mature and all that revenue is being scored under the solutions bucket. So that certainly has been helpful. But product businesses are doing also extremely well. We think the trend will remain very similar in 2026. We have guided mid-single digit. And the question I'm most frequently asked is that we have 5 quarters in a row we have been growing 7% or 8%, why don't you guide the same? That guide will assume that we keep taking share all 4 quarters also in '26. Can we do that? Answer is yes. But should we guide it? That will look a little lot of hubris that we think we can do it forever. So we're being more cautious. I think competition is smart. They are not waiting. So they will react to our share gain. But our efforts are not stopping. So we believe that we should maintain the same rate, and we'll report every quarter how we are performing there. But the solution product mix, we like 60-40 mix. 60% of our business is products, 40% of our business is solution. Within that 40%, 25% is services, 15% is projects. So our project business is actually very small. We are a pure-play products business. We sell products through channels. That's what our business model is. We have stayed honest to that business model for 6 years now, and it really works for us. So we don't want to change it.

Got it. So I already asked you about projects in the beginning. So when I look at process automation and technology, UOP is going to be a big portion of that. And I think you have good experience with UOP over the years. You've led that business back in the day. So why is it so lumpy these days? And you made a lot of acquisitions that are kind of around it, Johnson Matthey, Sundyne, Air Products. How can they help balance out the lumpiness, maybe?

So I think there are 2 parts of the story there. If you see the process technology business, the projects by default play out at this time of cycle of the project. So we have a good backlog, and we do expect good revenue from projects portion of that business in the second half of the year because we actually have the backlog. So we don't have to guesstimate anything there. The lumpiness is primarily driven by the timing of the catalyst demand. Now catalyst demand is not linear. It depends upon the production rate of our customer. So if you are producing at a lesser rate, you need it later. And that drives lumpiness, and it's a very high-margin product. So in good time, we sold lots of it. So everybody is happy. And when there's overcapacity at this point, customers are running at a lower capacity, so they are moving the demand to the right. And that essentially is creating a certain level of lumpiness because it's not linear. It's not if you bought in January '23, you will buy in January '26. It could be July, it could be September, depending on your production rates. We know about it. It's not a surprise factor for us. So I think a good thing always to look at process technology business as more linear on an annualized basis and not on a quarterly basis. It's just the nature of the business. There's nothing right or wrong about it. We know on an annualized basis, how will the business do. But within a quarter, you can have a lot of movement within 90 days period, but less movement within the 365 days window.

That's helpful. And then maybe focusing on your new Industrial Automation business. I mean you said in the beginning of this conversation, right, that what's left is basically sensing and measurement, right? So we understand that. But I think you can focus on improved execution, pricing, NPI within the portfolio there. So I think you said it's your biggest margin opportunity in '26. So maybe talk about that. And why would the business be down low single digit to flat if you have all these self-help initiatives that are going on?

So the business growth or the way we have guided flat to minus low single is more driven by end market drivers, primarily flattish Europe and lack of growth in China. The business is doing very well in North America, but that goodness gets offset by other factors. What we are doing well in '26 is given that we know the end state of this business, now we have a firm position in Industrial Automation. We want to be sensing and measurement business in industrial. Given we have a clarity, we can do structural changes on our fixed cost, which we started doing in '26, which is becoming basis of margin expansion in Industrial Automation. And I think that will remain optionality for next couple of years because now we are rowing the boat in a certain direction. Earlier, our boat was more around trying to find our direction, but we know we want to go this direction. So our new product strategy is working very well, as I mentioned before. There's no reason we can't do what we do in Building Automation and Industrial Automation. It's a very similar business model. We sell product through channels. It happens to be industrial products. There happens to be a building products. But fundamentally, it's a product which measures something. It has software. We understand the business model extremely well. And our confidence factor that business will perform well in 2027 is very high.

It's helpful, Vimal. So maybe just moving to aerospace for a minute, like if you could just break down your high single-digit growth a little bit more. I think you talked already about defense leading then commercially OE, then aftermarket. But aftermarket, as you know, has been strong for a while. So why can't it still be strong? And when I think about defense, are there multiple drivers for both the U.S. and international. So they're about even? Like how do you think about that?

So between the 3 segments, near term, the defense is the strongest growth. It will be high single to low double-digit growth in 2026. I think demand remains very strong, both in U.S. and international. Business jet, which is a big part of Honeywell portfolio is growing low to mid-single digit, too. So we do super midsize because within the BGA, there are multiple segments. So we primarily play in super midsize, which is working well for us. And then commercial aftermarket is growing mid- to high single. It's more getting synchronized with the overall growth. So overall, when you do the math, defense is more favorable in '26 compared to other segments. And we do remain confident of delivering high single-digit growth this year. I think the only gating factor in such a high demand-driven end market is how much we can produce. And we have delivered 15 quarters in a row, double-digit volume growth. But for us to deliver our numbers, we have to do 15 more quarters the same. It just -- and the base keeps growing. If you produce 100, 100 became 110, then it became 120. 10% on 120, now it's 130. So supply chain has to keep producing more perpetually forever. So that's only one variable in this business if -- how we develop our supply base, overarching supply chain capabilities. And I think that's what the Aero team is singularly focused on all the time.

But Vimal, it does seem like the Aero team has begun to really get on top of the supply chain stuff. It's not done yet, obviously, but like it seems like it's -- because you've -- the sentiment from you guys has changed a bit from tariffs, I feel like, where you're now talking about margin expansion visibly in that business. So it feels like it's that. It's CAES getting higher margin. Anything else I'm missing like [indiscernible] change?

Essentially, the -- there were some -- '25 was a bit of a unique here because we were passing through an acquisition. So CAES acquisition had an integration-related cost. We also had a unique year of tariff showing up on April, and many Aero contracts do not allow tariffs pass-through. So we have to just deal with that. OE mix was still unfavorable to a certain degree. So I think if you pass back to [ not '25 to road to '26 ], tariffs have become a new baseline. So we don't have to deal with that headwind anymore. CAES acquisition has largely been integrated. Supply chain-related investments have stabilized. So all factors are more favorable for margin expansion. The only variable actually for us, whether margin expansion will be the low end of our guide of 20% or high end of our guide 50%, is a mix of the shipments. What shipments go to which customer, which is driven by supply chain constraint. If we knew that finally, we can actually make it very, very precise. That variable is really the determining factor that whether it be the upper end of the guide or be on the lower end of the guide. But the fundamentals are very, very robust for us to drive margin expansion because the headwinds are behind us.

And just the commercial OE sort of customer agreements that you're sort of negotiating, can they be another opportunity for pricing and margin uplift?

Longer term, yes, not in 2026. I think these contracts are long-term contracts, and they will play out for aerospace margin expansion from '27 onward. They are -- this is not one contract. These are multiple. These are not commercial OE. There are some in business jets, too. And as these contracts due for renewal, that provides us an opportunity to mitigate some of the past cost of inflation, how do we recover that, pricing opportunities. So overarching basis, yes, that's an additional tailwind, which will start playing into the business from '27 onwards.

And just quickly on free cash flow. Like when you think about sort of 90% free cash flow conversion ex pension, 14% free cash flow margin, those are good metrics. But like as you become CEO of the RemainCo, do you think about -- is there anything stopping you from 100%?

I mean if you -- answer is no, but we have to remember that we have a noncash pension income as part of our earnings. We haven't really decided that it will be included or excluded. That's still in work in progress. But excluding pension income, I think high 90s will be the new baseline for RemainCo. There's no reason -- you're absolutely right. We are a light capital business. So pretty much all income flows into cash. The working capital investments are flattish. We maintain our inventory quite well. So yes, our conversion rate should be very robust moving forward.

Last question, Vimal. I ask you this every year. What are the top 2 or 3 innovations and structural changes affecting your company over the next 5 years? And are there any emerging industry trends that are perhaps overlooked in the current discourse?

I mean I think we talked about the structural changes to me are 3 from business I lead. First, we talked about inflation, so I don't have to. I think structural change. And what we need to think about is how does the industry deal with 3% inflation perpetually, right? It's going to be -- you have to be just not wash it over. We just need to be more thoughtful about it. I think geopolitical circumstances are certainly something to be dealt by a global company. Our RemainCo has 60% non-U.S. revenue. So we certainly need to stand up in these and make sure that we are local for local in every part of the world. And the biggest opportunity upside for us is how automation business will become an autonomous build, more towards autonomy. And I really feel excited that the timing is very good because automation systems create a lot of data, but their data is used to solve a defined problem. It's a rule-based system. How do you make a rule-based system to a gold-based system is going to be autonomy. So every time I look at a Waymo car, it reminds me that if a car could be autonomous, a building can be autonomous, an industrial plant can be autonomous. They have really solved a very hard problem. So we can do the same thing for the sectors we serve, and that creates a big opportunity for the RemainCo to transform the sector we operate in and create opportunity for our shareholders.

Well, we're going to talk about that very soon, Vimal. So appreciate the time. Thank you.

Thank you very much.