Flex Ltd. (FLEX) Earnings Call Transcript & Summary

Okay. Great. Thank you, everybody, for joining us. My name is Mark Delaney, and I have the pleasure of covering Flex for Goldman Sachs. I'm very pleased to have with me today, Michael Hartung, the Chief Commercial Officer of Flex as well as Michelle Simmons, who heads up the IR function. Thank you both for joining.

Glad to be here.

I thought we could start off with some questions by end markets. And certainly, one of the bigger ones we've been hearing a lot this week at the Communacopia and Technology Conference is on the data center market. For Flex last year, a $4.8 billion business. Now nearly 20% of your revenue, I think $3.5 billion of that $4.8 billion was tied to assembly and then $1.3 billion in power. Maybe you can speak a bit more to the breadth of your data center customer base. What's driving it to that size? And what do you think is positioning Flex well in the data center sector?

For sure. So first, maybe an update on the business itself. So we talked about, to your point earlier, starting the year off at that $4.8 billion number, and we're going to end this year at $6.5 billion. So growing that business about 35% year-over-year on the heels of a growth rate of 50% the prior year. So still really strong growth in a really exciting market. Also would point out that of that $4.8 billion, that $1.3 billion was power and the balance was the integration business. And of that 35% growth rate, we think the power business will probably grow a little bit better than 35% and the integration business may be a little bit lower, but also the sizes of the businesses are different, too. In terms of the composition of the customer base and so on, probably easiest to answer first by reiterating the nature of the portfolio. So as you may or may not know, Flex provides both products and services to the data center. On the product side, we have 2 different power businesses that leverage Flex created IP to bring products to market. The first business is around critical power. And in that space, we tend to compete against some well-known companies like the ABBs, Eaton's, Schneider's, Vertiv's, that critical power infrastructure, we provide things like low-voltage switchgear, medium-voltage switchgear, busways, bus bar, all the power infrastructure that goes into the 4 walls of the data center itself. In addition to that, we have a second product business that's really focused around embedded power. And that business takes that power from the infrastructure into the board level and at the rack level. So both critical power and embedded power from a product perspective. And then on the other side, we have a services business where we focus on IT rack integration and tend to vertically integrate those racks. So we can assemble that rack, we can also vertically integrate that with our own sheet metal fabrication capability with our Coreworks product line, which is essentially electronic and mechanical components. We can even bring the aforementioned embedded power group into that product as well. So really well diversified across that grid-to-chip sort of spectrum, which gives us access to 3 distinct categories of customers, which I think is also relatively unique. So obviously, hyperscalers is a big part of the business. The good news in that space for us is we don't have a heavy concentration. We have multiple hyperscaler relationships. And also within those hyperscaler relationships, we tend to sell multiple products and services. So great diversity amongst hyperscalers and in the portfolio that we offer. Unique and new to us is that as we've built this critical power space, we've also created engagements with the largest colos. So hyperscalers and colos and then obviously, the silicon providers. And we engage the silicon providers in 2 ways. One, when those silicon providers develop their next-generation ASICs or GPUs or CPUs, they need to address power consumption. And so our embedded power group jointly designs a lot of those applications, gives us access into that space. And then we also work with those same companies to build the boards, integrate the racks and the like.

Michael, in your role as Chief Commercial Officer, you're engaged with a lot of the customers have a lot of conversations. As you look at the things that are driving that, mid-30% range growth rate for fiscal '26 in the data center business. Talk a little bit more around the breadth of that and to what extent you're starting to see either Neoclouds customers or even sovereign, enterprises starting to do some AI data center. Where is that strength coming from?

Yes. The interesting part is it tends to continue to broaden. And it depends on the product or service offering that we're providing. So when you think about the IT rack integration space, certainly, the hyperscalers drive the vast majority of that demand. But we're seeing more and more applications at the enterprise level, somewhat due to the fact that we recently announced an acquisition of JetCool, which is our liquid cooling platform. And in that space, it's actually given us a nice entree into more enterprise type of relationships that prior, we really didn't have a lot of exposure to. So the breadth of the portfolio really does help broaden that. Now we also have a really intentional effort to get beyond hyperscalers to talk about and work with the NEO scalers as well and seeing more opportunities with them because they tend to want more integrated turnkey solutions across a broad spectrum of products and services. And so because we offer this really unique portfolio that goes from critical power to embedded power to IT racks integration, we can provide more through one supplier versus them having to go to multiple people.

Very helpful. The company recently announced a warrant agreement with Amazon. Help us better understand that, if you could, please. Why now? What does it add to the financial outlook for your data center business?

Yes. I wish I could share a ton of detail around that because I really believe it's an exciting relationship that we've built here with Amazon. We're really limited on what we can say, and I'll defer to Michelle on maybe some of the specifics. But some of the key takeaways that I would think about. One, I would see it as a really strong endorsement of what the opportunity is with Amazon. I would reiterate that it's for the entirety of Amazon, not any one particular business, and it applies to our entire products and services portfolio. Thirdly, I'd say it provides incentive to grow the business. So even though I can't get into the specifics of what we're going to win, when it's going to come, there's certainly incentive to make those things happen. And what I would say is I would look back to really what we said 2 years ago at Investor Day when I think we first said we wanted to grow this business about 20% CAGR for the next 4 or 5 years, ended up doing 50% last year, 35% this year. This agreement wasn't in place when we made those commitments. So we're really excited about the possibilities.

Michelle, do you want to add anything?

No. I think I would just say it's a long-term partnership that we're super excited about.

Do you think this kind of structure could be something you would do with other hyperscalers?

Well, I would say that we have a wide variety of models in place with all of our customers today. Some customers prefer to really do more of a turnkey relationship and put that burden on Flex. Some customers prefer an asset-light model where they do a lot of that investment. We're open to any variety of financial relationship that makes sense that provides the return that we expect and the value to the customer. So the agreement doesn't prevent us from doing anything, whether it catches on or comes into play with other relationships, really hard to say.

The very strong growth in the data center market is also attracting competition. So to what extent are you seeing more competitors come into that space? If so, what does that mean for margins and how you manage your margin structure for the data center space?

Yes, great question. And I would first -- I would start with where we're tending to focus on our because it's a really broad market landscape that we're talking about here. And the first distinction I would make is that we tend to focus on custom applications, not standard platforms. And that's important because the standard platform space really tends to be dominated by ODMs. And we just don't see that market opportunity being margin rich. It's high volume, tends to be lower margin. It tends to be movable more easily. So we tend to focus on custom. And because of that, we tend to focus on, for instance, hyperscalers that bring to market their own silicon, right? They tend to require more custom applications. I'd also say that next level of distinction is from a competitive standpoint is how we compete today. So think about EMS. So we've been in the EMS industry for 50 years, I think, and we've built this network of 150,000 people and over 100 different facilities in 35 or more different countries over that 50-year time period, and that's the foundation of what we do. But we're very different today from EMS in the data center. How is that? Well, look, a lot of EMS providers can integrate a rack. That's not necessarily differentiated. We could argue that we do it at scale that maybe some others don't. But where we become differentiated compared to an EMS provider is we can vertically integrate that rack with our own sheet metal. We can vertically integrate that rack with our own embedded power with our own Coreworks product line. What EMS providers don't have is a power product portfolio at all, right? So even if an EMS provider says they're in the power business, they're build to print. They're building someone else's design. We're actually investing in and creating Flex IP to bring products to market. So we can compete in just EMS, but we're very different across the portfolio. From a power perspective, kind of a similar situation. In the critical power space, I mentioned that we compete with some pretty big names, the ABBs, Vertivs of the world. But those critical power providers don't provide embedded power solutions. And they don't provide IT rack integration at the scale in which we do. Similarly with embedded power, we tend to compete with companies like Delta, right? Delta is a good company in the embedded power space, good competitor, market leader. They don't provide critical power solutions. And they don't provide IT rack integration at maybe the level and the scale that we do. So I think we've got 3 different categories of opportunity. We're able to compete in each category, but the more we're able to provide an integrated solution that requires 2 or more, we become really different than the other alternatives out there.

You just mentioned the importance of the whole portfolio and that full solution that you can bring to the customer set. You've done some deals to get to that level. You already mentioned JetCool. You've done Crown Technical. As you think about what you have in the portfolio today, how comfortable are you with this current product capabilities that you have in-house? And do you look to add more either with R&D or inorganic actions?

Yes. We -- as you've seen, we've been pretty active in the market, and we're deploying capital against those things that provide the greatest return. We'll continue -- if you look back, you'll continue to see the same level of discipline, whether that's through buybacks, whether that's through organic growth or inorganic growth. I'd say when you think about the question around the data center itself, we've mapped out the ecosystem there. We've come out and said, we think our products and services portfolio covers about 80% of the available spend. So then the question becomes, how do you penetrate that 80% to a greater extent? And how do you get after the 20% or do you even want to? And I'd say that -- we think about pursuing that through 3 different lenses. The first lens is make. Is there a way that organically, we can create that capacity and that capability on our own. And I think to a certain degree, the answer is yes. And you've seen some announcements for us where we've announced a new factory in Dallas to support our critical power business. We have announced a new factory in Poland, a new factory in the U.K., all to develop and support the growth in those areas. In terms of acquisitions, we'll continue to look for opportunities to scale that. If it provides a capacity benefit, it provides a capability benefit, we'll be thoughtful if it provides us with the right return. And then the other area, the third area that we're mindful of is you don't necessarily have to make or buy everything. right? And partnerships are going to become a really more important part of that overall portfolio as customers are looking for a one-stop shop for an integrated solution. We're looking at it from a make-buy and partnership perspective.

One of the near adjacencies for your power product set is the utility market. I think it's an area that Revathi and Chris Butler, who are really important new leaders in your company have a lot of experience with. So as you think about the products you have, some of the management knowledge of some of these close adjacent markets like utility and grid, to what extent do you think you can expand into some of these areas, especially now that you have something like Crown Technical?

Yes. We're excited about what the options are here. And when you think about the Crown deal, first, you have to step back maybe earlier in time and talk about the Anord Mardix deal, right? So the acquisition of our Anord Mardix business really provide us an entree into the critical power space. It was focused, however, in Europe primarily. And so low-voltage switchgear, power pods, busways, bus bars really focused on the European market and focused around data centers. When you think about the Crown piece, we've got a number of things from that deal. One was we got medium-voltage switchgear to the portfolio. Two, we had an ability to extend that offering of power pods from Europe into the U.S. And then thirdly, we had access into the utility market. And so I would expect us to continue to pursue that opportunity, especially when you consider that power is the constraining item when it comes to the growth of the data center. So you'll see us support both the data center side and the utility side.

Maybe shift gears a little bit to the automotive market. It was one of the areas that the company articulated as being a key growth driver for it at your last Investor Day. I think as of that point, you were expecting that 10% CAGR. There's obviously been some volatility in terms of the policy environment, macroeconomic conditions. Can you level set us on what you've been seeing in the auto space and how investors should think about growth here going forward?

Yes. The first is the obvious statement. It's a far different world today than it was when we made that commitment for sure. And I don't think we're alone in that. So whether it be macro, whether it be policy, it's been a pretty tough road for the automotive business, and we've been upfront about that. Nothing new since our earnings, but still in the same challenging spot that maybe it was. Maybe the more important piece of that is where do we go from here in that space. And a couple of developments I would point to. One is, first, there's been a lot of talk about the false start in the EV and now the pivot back into hybrid or ICE type applications. And so the first thing that we like to reiterate is that we've come to market with a power platform that is agnostic to either hybrid or EV. So once the companies that we support finish the retooling of their factories from EV back to hybrid, we're in a really good position to sell our power platform into that space. The second area I'd say is from an automotive standpoint, consumers in that space are transitioning from buying automobiles based on, say, mechanical performance or engine performance and user experience. And right now, the big impetus for our customers is around the software-defined vehicle. right? The software-defined vehicle really suggests that software is becoming the core competency, the next secret sauce of the automotive makers. And that's created an opportunity from our perspective. Because up to this point in time, think about the market being served in 2 ways. One, from a build-to-print standpoint, where EMS played build-to-print whatever the automotive maker wanted, that's what the EMS provider to build. On the other end of the spectrum, you have the Tier 1s. And the Tier 1s have been continually investing in the software capability to marry with the hardware capability to provide a complete product to the automotive makers. Well, that worked for a number of years until software became the next secret sauce. And so that's created a bit of a problem because I don't know that Tier 1s are too enthusiastic about not selling software to go with their hardware given the investments that they've made. So we decided to grab that middle ground. We don't intend to be a software provider in the space, but we also don't intend to be purely EMS. We've actually created our own product platforms, centralized compute to support the trend towards software-defined vehicles. So our automotive customers can take that platform, apply their own software and go to market with their own customized solution.

You mentioned power. You just brought up compute. Any other key content drivers for you in the automotive space, anything around ADAS sensors, actuators, other important drivers we should have in mind?

Well, I'd say that what's driving that compute platform, that compute platform is going to be centralized, and it's going to support all of those adjacencies that you talked about. One of the big transformations taking place in automotive today is it used to be a very mechanically driven distributed architecture and how an automotive -- and how a car was built. Well, what the automakers have learned is that there's a better, faster way to do this when you look at the pure-play EV providers who have come to market without the constraints of how we used to build cars, right? They're native online computer-driven sort of vehicles, and they're very centralized and very simplistic. So automakers are transitioning hard into that way of thinking. So having this ruggedized centralized compute capability is going to play into a lot of those applications.

You mentioned something interesting I wanted to follow up on about having product capabilities in the auto space. And we were just talking about the data center market and having products and the margin profile of selling that product on your own to some of those customers. As you think about autos, is that same kind of an opportunity? Or do you need more you've got some reference designs that can speed development if customers choose, but still a little bit more in line with the past model?

Yes. I'd say it's closer to the reference design end of the spectrum. On the power side, there's an opportunity to have that more be a product offering because that's what automakers are looking for. They're looking for a more turnkey solution on the power side because they don't see a DC to DC converter as being a core competency like they do on the software side. So in the compute space, you'll continue to see that be a platform where they can integrate our hardware with their software, but it won't be a complete product platform because, again, software-defined vehicle is their new secret sauce.

And where are you seeing the most traction with customers in the automotive space? I think historically, Flex has skewed a little bit more to some of the traditional auto OEMs for revenue in this area. Are you having any momentum with some of the Chinese auto OEMs or EV pure plays? Just help us understand where you stand at that.

Yes. What we've said is still true today, and that is our largest customer base is North American focused, followed by Europe and then followed by China. Now we state we're engaged in China for a couple of reasons. One, it's hard to argue that there's not a massive amount of innovation taking place in that marketplace, especially around EV. So we want to participate in that innovation for sure. Most of our business, if not all of it, is in China for China when it comes to that geography. We are being pretty thoughtful though, because in China, you also have hundreds of automotive OEMs, and there will be consolidation. Who the winners are beyond BYD is probably hard to predict. But for sure, we know there will be consolidation in that space. We want to be thoughtful about it. We want to have exposure to it, but we'll be thoughtful about China. I'd say that North America and Europe continue to be the vast majority of our opportunity. It's both ICE and whatever EV is the choice of the marketplace, whether it be EV purely played or hybrid. I would say, though, that more of our opportunity is starting to look less like pure-play EMS and more around this platform plays around power and compute.

Very helpful. Maybe talk about a couple of other end markets that Flex has exposure to. I wanted to touch on networking. Talk about what you're seeing from an end market perspective, but also some of the share opportunity because I think you alluded to some potential share gains on your last earnings call.

Yes. Networking has been an area of improvement for us, which is pretty exciting. And first step back and if you think about our data center business, we talked about that business being about a $6.5 billion business. by the end of the year. And we've narrowly defined what goes in that data center business. It's hyperscalers, it's colos and it's silicon providers. It's not networking, right? Now what we all know is that more and more networking has exposure to AI-driven demand. And so a couple of things are happening in that space. One is the complexity of the programs that are being asked to manufacture is continuing to increase and the places that they're asking for these things to be manufactured is shifting. So for years and years and years, low-cost supply chains were put in place in Asia, in China that were hardened over 20, 25 years. Now they want to take more complex technologies and transition from areas of, call it, higher risk into areas of lower risk that are closer to point of use. And so we're shifting that production. As we're going up the complexity curve, we're shifting that production to areas like Mexico, Europe and India.

And is that just kind of traditional switching and routing or optical or both?

The whole spectrum. So we're a large manufacturer of high-end switchgear. We're not a product company when it comes to optical, but we're a large manufacturer of optical equipment for our customers. And so when you think about that spectrum of products, I would say what shifted is an ongoing trend towards the higher end of the complexity spectrum and more regionalized manufacturing of those applications.

Okay. Very helpful. I want to talk on health solutions. That market has been a little flatter over the last few years, especially post some of the work you all did for COVID and some of the respiratory products have been in the $2.6 billion to $2.7 billion range. You talked about an expectation for some recovery in medical equipment toward the end of this fiscal year. What drives that? And what should investors think about growth over the medium to longer term in medical?

Yes. I'd say when you think about the health solutions business, think about it through 3 different lenses. We've got our medical equipment business, our devices business and our drug delivery business. And if you think about sort of the performance of the recent past and the expected growth of the future, it's probably easiest to say that the softness in the equipment space is being offset by strength in the devices space with hope on the horizon for drug delivery. Let me peel that apart a little bit, okay? So first, on the equipment side, I mean, CapEx spending is still pretty tough in the capital equipment space. Interest rates are still high. I don't think that spending is recovering in a really rapid rate from that perspective. I'd also say that OEMs in that space are facing increasing competition from Chinese medical equipment companies that are doing very well in that space. So we're not seeing a big bounce in that space. We are seeing strength in our device business. So we are a big provider into the CGM space, the continuous glucose monitoring space. So that business is continuing to grow for us. We do expect that to continue. So real simply, softness in equipment offset by strength in devices. That leaves us with drug delivery. So we've announced some recent wins in that space in the GLP space. And we're excited about that. But also wins in that space take a few years to ramp to production. So that's why I say on the horizon, we expect growth in that space. So that's how I'd characterize the health solutions business for us.

But it is one, I think, has the potential to grow at least mid-single digits, if not even something a little bit better than that as you think over 3, 5, 10 years.

Yes. And we haven't updated our guidance on that. I think we have an Investor Day coming up in May, where we'll refresh the outlook, but we would certainly have a lot more optimism around the health solutions space than maybe some of the other businesses.

Okay. You mentioned services previously in our conversation. You spoke on some of the capabilities Flex brings to that area around metal vending, but I think it's a broader capability set that Flex brings. Talk a bit more on what you do in services.

Sure. So I'll just remind everyone to step back first, EMS was what we had been known for that we've built over the past 50 years. Now we're a products and services business. So think about that lens first. Take away the product business, right? Now you're left with services. Within services, you have 2 different areas. One is traditional EMS and the rest is the value-added services piece that you're talking about. So what does that mean? Well, it starts with the front end. So we've created a design capability separate from the design and engineering that's focused on creating Flex IP and products. This is to bring products to market for manufacturability, bring products to market faster, more reliably, more cost effectively based on what we know about the manufacturing environment. That's one piece. Secondly, think about the products that we're manufacturing. We vertically integrate that with things like sheet metal. We vertically integrate with plastics with our Coreworks product line. So that's another piece of it is the vertical integration. And then the third piece is around aftermarket. And so we have a global services business that operates post production that's really operating at scale already that we don't talk a lot about, but it's operating in over 25 factories today, and it's really 2 different categories of services. One is on the forward side where we provide value-added fulfillment capabilities. So we can do late-stage configuration, and we can land product on time, on cost, on quality, either to another business or even to a home address, depending on the business that we're providing that service to. Alternatively, once it goes out, some of them might have to come back. And so through our circular economy suite of services, we can repair, refurbish or recycle those products that we put into the market and either repurpose those for resale, but for sure, limit the amount that goes into landfill.

And I think the services business, if I'm not mistaken, has pretty attractive margins.

So the services business, I think, is important for a number of reasons. The first of which is it does operate at a margin profile that's greater than the company average. But the second is it provides a more sticky relationship with our customers. So we talk about competition a lot. Well, you have to differentiate to earn the right to keep a customer. And so the more products and services we can do with each relationship, the stickier we get. And so much easier to keep a customer if you're not just, say, integrating a product, but you're vertically integrating it, you're fulfilling it, you're taking it back for service, much harder to disassociate that relationship than just pure building.

I wanted to talk a bit more on manufacturing and your global footprint. Flex has, I think, about 100 sites across something like 30 countries. You already spoke around the networking and optical space and seeing some of the customers want to do more close to where they're ultimately shipping their product. But talk a bit more on what you're seeing from geographic trends and to what extent, especially with tariffs in place, you're seeing more of a shift to your Mexico and U.S. sites?

Yes. I think the first principle in all this is expect the unexpected, right? And although that sounds like that's something you can't plan for, it's actually a really important principle for which you design your supply chain, because you have to design for unpredictability, mean long gone are the days where you can rely on one single site in location to optimize for cost. You now have to optimize for business continuity as much as cost and for customer experience, bringing production closer -- as close to the customer as you possibly can and to your point, to mitigate whatever the trade policy ends up being whenever it actually happens. So I'd say that you've seen the investments reflect this trend for us. And I'd say these trends have been occurring over the past 5 years. This isn't a recent phenomenon. I'd say it dates back, I'll call it, Trump 1.0 for the first administration when trade wars really started, followed by a global pandemic. followed by massive material constraints, followed by real wars in multiple geographies and now, call it, Trump 2.0. So we've been in this environment shifting the supply chain for a number of years now. And what it's resulted in is a growing capability and capacity in Mexico. Now that we've had the benefit of hosting down in Mexico to show you that capability before, and that's been a great success for us. We're also investing considerably in the U.S. today across multiple businesses. We talked about our investment in Dallas. We've talked about our investments in South Carolina, among other places. So the U.S. footprint, North America is one of our fastest-growing footprint in the portfolio today. You look at Europe, we talked about investments in the U.K., talked about investments in Poland, another recipient of capital when it comes to the portfolio. And then thirdly, one we haven't talked about today is India. And India is still a really important capability for us, depending on what that opportunity might be. So from our perspective, we're going from this low-cost single-site solution into a more regional operation and climbing the complexity curve at the same time.

How important is automation in your manufacturing, especially as you do more in the United States, do you need to do incremental automation? And if so, are you seeing anything interesting on the technology front that is allowing for that?

The more you shift into these new regions and geographies, the more automation will play a role. First, you have to talk about the data because everyone really is interested in how you're utilizing AI to optimize your efficiency and your productivity. And the first thing you have to sort of address is the data. And right now, that's one of the biggest challenges to applying AI to a manufacturing environment. All these different equipment platforms don't provide data in the same format to make it usable and leverageable. So one area that we're really focused on is harmonizing the data from which we can apply AI. While we're doing that in parallel, we're obviously not waiting for that. We've invested in our own Flextronics standard equipment platform. So right now, today, inside of Flex, we have a standard platform that applies to -- last I counted over a dozen formerly manual processes in our manufacturing environment. So now we can automate with our own platform on any factory to get to that automation more quickly and more cost effectively and get the data in a harmonized way so that we can then apply AI. So doing a lot of work around data harmonization, a lot of work around deploying the standard automation footprint. And I would say the more that regionalization impetus takes hold, the more we're going to rely on automation.

Are humanoids something that you guys think you'll be using? And if so, is that next couple of years or 5-plus years?

For sure, everyone has to contemplate the use of humanoids in a factory. I don't see that as being a near-term opportunity at present. But I always get a little reluctant to doubt the rate of innovation that takes place. And so I'd never say never. I'm not locked in on a time frame. But near term, we're focused on data and standard platforms. And if humanoids become a good use of capital, we'll be one of the first implementers of that.

No, always interesting to see all the things you guys are working on given all the different products and regions that you manufacture in. I did want to ask another one on end markets. We spoke a lot already around some of the different end market trends, especially over a medium- to long-term perspective. But coming out of the most recent earnings report, has anything surprised you in terms of what you're seeing in terms of market conditions?

No, I don't think there's any -- yes, so first, I don't see any real change since earnings. So anything I say isn't intended to communicate a change at all. I would say that the good news is we're seeing continued strength in data center. right? We're seeing the power businesses perform really well. The IT rack integration business is running really well. We've seen the ongoing sort of challenges that we talked about in automotive. We've talked about renewables. The thing about our portfolio is it's built to withstand the variances that happen by market. I've been in the company for 20 years now. I don't think I've ever been here at a time when every single business is operating at peak performance at any given point in time. But I think it's periods like this where the macro is, relatively speaking, a challenging place to be and still being able to put numbers on the board, improving margin, improving profit dollars, delivering value to our shareholders is in some way, shape or form, a result of the strength of that portfolio.

End on a margin question. I mean, because you guys are at 6% margins, even with some of the tariffs and roughly double what you would have been doing 5 to 10 years ago. So it's been a remarkable journey. Is 6% sort of a peak margin level as you guys think about the profitability potential of this business? Or is this something that you can build off over the longer term?

Yes. So I can't refresh our guidance yet on what the long-term views of the margin is, thrilled that we're able to get to that 6% level almost a year in advance of when we thought we could. And if you look back, you probably get the answer to where we're going in the future. First, we've been very intense about shifting the portfolio to higher-margin businesses. When I was running the Agility Solutions business, we talked about really methodically taking down our consumer business and replacing that with higher value, more stable business. And sure enough, if you look back, we took our Consumer Devices business down by $2 billion. We replaced that with high-calorie data center business and improved margins from a 2 handle to a 6 handle over that time period, right? You've seen our reliability business even in the last year, improve to that 6%. So we have both segments operating at 6%. When you look at the growth rates in our higher-margin business of data center and you look at how that portfolio could shift, you'll naturally get some blended margin improvement as a result of that expectation.

Well, it's been really great to see, and we look forward to what you have coming ahead. Unfortunately, we are out of time. So we'll have to end it there. Michelle, Michael, really appreciate you joining us.

Real pleasure. Thanks for having us.

Thank you.