Flex Ltd. (FLEX) Earnings Call Transcript & Summary

Well, thank you all for joining us today. Excited to be spending time with you here, and we're fortunate to have our guest today for our conversation, the Chief Executive Officer of Flex Limited, Revathi...

Advaithi...

Advaithi, sorry. I know, I panicked right at the second. So maybe for those not familiar with the Flex story yet, why don't we start there and kind of level set the audience on the background of the company.

Okay. Great. So Flex is a contract manufacturing company, one of the world's largest, around $26 billion in revenue. And a good way to think about Flex is we're kind of the name behind the brands. We make something for everyone. I would say we report in 6 business units, but think about it as participating in 5 major end markets. Two of the business units are in the consumer end market called lifestyle and consumer devices. One of our biggest end markets is what we call CEC, which is cloud enterprise communication. We're big in the health care space, particularly around devices. Big in automotive and the last one being industrial. So those are the major end markets we're in. Today, we talk a lot about Flex as also being a data center AI infrastructure player, which means that if you think about the $26 billion, 75% of the company is in the traditional contract manufacturing space, 25% of the company is in the AI data center utility space and growing at 35%. So it's one of our highest growth areas and also an area where the company has become more of a product company versus a contract manufacturing company. And in this space, we do compute, cooling and power as individual components and as modular integrated components for the data center play. We are in 30 countries across the world, 100 different manufacturing sites, around 140,000 people. That's a little bit about Flex.

Okay. And I do want to just give a little background on yourself because you had joined Flex just prior to the pandemic. So maybe you can share a little of your background and also maybe what you've seen change at Flex in your time in the scene.

So I joined Flex in February of 2019. So I'm in my seventh year. And I came -- most of my background is with two companies, Eaton and Honeywell, so mainly in industrial end markets, everything from oil and gas to construction to electrical, most of my career with the electrical or energy space. And when I left Eaton, I was running their electrical business globally and before I came and joined Flex in 2019. And a little bit about the kind of space itself when I joined, I obviously came from the industrial space, diversified industrial, didn't know much about contract manufacturing, but thought that it felt like a space that was ripe for change, and I wanted to do something different than traditional industrial. And three major players, felt like financially, the space hadn't done well. And I felt like, hey, if 1 of the 3 kind of has certain more fiscal discipline, it will be a good thing for the industry. And so the focus in Flex was to say that we will run our portfolio in a more disciplined way, which means that we won't grow for the sake of growing. It was not all about growth and capacity expansion. It was going to be about growing with great cash flow, good margin improvement. So we exited many end markets, particularly around consumer end markets that we felt wasn't financially viable for a company like Flex. And we focused on difficult, harder to do things. And so if you think about the last 7 years, we went from kind of 3% operating margin to 6% operating margin today. Managed the portfolio extremely well by really focusing on mix. We also spun off another company called Nextracker, which today is around a $13 billion, $14 billion market cap company on its own. So really a lot around portfolio and running it well, also drove tremendous productivity for the company. So through the pandemic, through the supply chain crisis, through all the geopolitics of war and macroeconomic changes that has happened as a result of tariffs, we've really continued to perform extremely well as a company. And our shareholder return has been great. We've been rewarded well by our shareholders. And so it's been a good story so far.

It sure has been. So starting with the data center business, you had reported, I believe, $4.8 billion in revenue for fiscal '25, about 20% of the revenue mix. What is that comprised of? And where -- why do you feel you're well positioned there?

So -- and that is going to be -- we have also reported that, that's going to be $6.5 billion this fiscal year and is growing at 35%, that same business. So let me talk about what sits in that portfolio. What sits in that portfolio is three things at a high level: compute integration, cooling and power. And let me talk about each one and what differentiates us. So in compute integration, we are like probably any traditional contract manufacturing. We will integrate server, storage, all of that together for a hyperscaler and test it fully and deliver it to their data center. The other thing that we do like industrials is we will be cutting your metal, building enclosures, doing all of that. In the cooling space, we mainly do liquid cooling. So we'll make the cold plate that goes on your chip, and then we'll make all the cooling that goes around the rack, including cooling distribution units. And then we do power. We do power within the rack, so what powers the chip itself, and we do power outside of the rack all the way up to the utility. So if you think about this for a second, companies will do bits and pieces of all of these things, but very few do all of them in an integrated way. And with power being such a significant part of what's required to power the chip itself, the efficiency that's required for it, the heat that it generates and the ability to cool it down, we are one of the few companies that will integrate it end-to-end. And we may put it in a big kind of container, think of it as a pod and deliver a complete solution to your data center, too. So individual components or modular or complete infrastructure build-out in compute, in power and cooling is what we do.

Okay. And who -- how is the competitive environment in that data center space for you?

Yes. I would say for individual bits and pieces, there's somebody who does pieces of each of it, like embedded power, for example, which is the power that powers the chip itself, which is the big conversation these days, there's basically two companies, right, today that does it Flex and Delta. And then if you think about power outside of the rack, which is the typical electrical distribution stuff, it's the big electrical players. It's Eaton, Schneider, those kinds of people. If you think about cooling, very fragmented, a lot of small players could be a Boyd, could be a Vertiv, could be a Modine, that kind of cooling play. For compute integration, it could be somebody like Jabil. For racks and enclosures, it could be somebody like an nVent or a Rittal. So the bits and pieces of it, lots of different competitors do. The real question is, as compute gets power hungry and power generates a lot of heat that requires to be cooled and what data centers really want is the ability to get speed of installation, who are the players who are able to do fully integrated solutions end-to-end and deploy it with speed, I would say that really reduces the competitive landscape in a pretty significant way.

Absolutely. Well, you certainly have an edge there. And it seems like one of the edges is your power kind of that differentiates you. And you mentioned the power pods. Can you share a bit more on that and what exactly is included in the pods?

Yes. So I'd say just if you think about our power portfolio, I talked about we make the power module that powers the chip. Then we put together the power infrastructure that goes outside of the rack and in the data center, like low-voltage switchgear, medium-voltage switchgear, all of that, power distribution units, data bars, busbars and then we go all the way up to the utility. We call that kind of chip to grid. It looks like everybody calls it that these days. But the real thing, what people want is these pods. And what the pods could comprise of is think about everything I just described, thrown into a giant container, fully tested and integrated and dropped into your data center or outside your data center powered up and ready to go. And that's what these pods are. Traditionally, people were doing it in the oil and gas space. You saw that when we were bringing up a big oil field or an offshore site, becoming more and more of a thing for data centers. But what's very cool about that now is it's not just happening in power, it's also happening in compute and cooling. So people are also now deploying IT pods that they're calling, which are the same structure, but with fully integrated with a compute rack and cooling and power, all that thrown in. So the capability that we have in power pods, which is fully integrated electrical infrastructure now is also being deployed to kind of IT compute pods also. So the whole idea is, can you improve my schedule and with labor being such a significant issue to deploy these -- the data center infrastructure, it also reduces kind of your labor costs and makes it super-efficient. So that's what these pods are. Very huge complex things.

Sounds like it -- it sounds like you've got it figured out. So Flex acquired Crown Technical, a company with modular solutions and medium-duty switchgear products, some helpful, including a presence in the utility market. How large is Flex's utility business? And how fast do you expect that area to grow?

Yes. So the reason for buying a company called Crown Technical was one thing to note about the U.S. utility space is it's super fragmented, which means that you have to be very regional and present with utility players to be able to really have a play. And I've spent a long time in this space. And when we built this data center play for compute and power, one thing that I wanted to do was really migrate more to utility because, as you know, a lot of conversation about utility bringing power to data centers. And we also wanted to get in the U.S. this ability to build large power pods, which I told you utilities were already doing. We already had that capability in EMEA, but it isn't transferable. You need to have local capability and local regulatory approvals to do it. So Crown gave us the presence in the utility sector, but it also gave us the ability to expand those power pod capability into data centers that we didn't have in the U.S. And then it's just diversifying. I like the idea of constantly diversifying out of a certain space. So it was going from data centers to utilities and helped us kind of really step into that product portfolio a little bit more in terms of what we were looking to do.

Okay. And as you think about your broader power business, does Flex need to add capacity? Or are there any key supply chain constraints?

I would say capacity will be a constant thing for us. And because we are growing, we're growing at 35%. So that brings the need to add capacity. And so we're evaluating capacity all the time, but the capacity decisions that we make today will be for 2 years out. And so I'd say, yes, capacity will be a constant thing. We're looking to add more power and the capacity we have because testing of these products require quite a bit of power. So we're looking to add more of that. So I would say most of our additions are happening in North America and to some extent in Europe, where we just recently added capacity in Poland. So yes, we're constantly evaluating our capacity. But one way I think about it is globally, what are we doing with capacity. So we're shrinking in certain parts of the world, and we are adding in others.

Okay. And how are lead times trending for that?

I mean that's difficult to answer question because I would say for power products, lead times has been like 12 to 24 months for a long time. Low-voltage switchgear, medium-voltage switchgear has long lead times. Power pods is pretty significant lead times. But if it's things like data bars and busbars, smaller lead times, I'd say compute racks and compute integration is all kind of planned out. Like the things we're doing for next year, we already know now. So it's really not a lead time driven. It's a capacity-driven exercise. So yes, it depends based on the product. Our product breadth is so large that some things are a couple of years out, some things are the next few days.

Yes. Okay. Flex also acquired JetCool last year, a provider of cold plates and CDUs. Can you speak to how your liquid cooling portfolio is doing?

Yes. So the thinking around liquid cooling, again, was what I talked about, we have power that's generating heat, we needed to cool it down. So in JetCool, we do everything from cold plate. So the cold plate is what goes on the chip to cool the chip down. And it's super interesting. Think about this as the chip, there's a power module that's powering the chip that's generating all the heat. And then there's a cold plate that's going on it. Think about it like a giant sandwich, right? And that is cooling everything. And obviously, that cold plate is very high technology because you don't want liquid to be going on your very expensive chip. So we're making all of that. And from that, we do cooling technology, which is basically pipes and stuff like that, that's cooling the rack itself, going all the way to the CDU or the cooling distribution unit, which is sitting either next to the rack or at the end of the row of racks. And as these racks are becoming higher and higher power density, the cooling units are becoming higher and higher power density. So what we're deploying with these RCDUs are the smallest I'd say, square footage modular CDUs. So data centers don't have to use up a lot of space, making these giant distribution -- cooling distribution units. So with JetCool, we're able to deploy a very modular CDU that is very scalable and can go all the way from close to your rack to end the road to be able to cool the entire network of compute racks that are going in. And that's what we did with JetCool. So again, compute we do, cooling, we do and then power we do, we put it all together.

Okay. We've discussed data centers and utility. Are there other businesses? Or how are the other business lines doing as well?

Yes. So like I said in the beginning, we're -- in our contract manufacturing space, we do health care, automotive, industrial and consumer. So I'll talk about all of those here for a second. Health care space doing really well. The big focus for us there is medical devices. So we're one of the largest manufacturers of continuous glucose monitors. And as you know, the world has a diabetes epidemic. So obviously, that's a huge space for us. We invested in that many years ago. So we're very focused on the device space in the health care sector, and that's doing well. I would say in the industrial space, it's a tale of two cities. So we had a big renewables presence in residential renewable, and that saw a huge decline over the last few years. And that has kind of bottomed out. And where we're seeing investments is in infrastructure, industrial deployment, more around energy, making large-scale utility grade storage or inverters, those kinds of things. That's where we're seeing growth in the industrial space. Automotive is what I call it as finally some clarity for which platforms are going to get deployed moving forward. So somewhat bottomed out in terms of the automotive space itself for us, but with good clarity of which platforms will get invested in. So I feel like there's growth from where we are at today. We are platform agnostic in automotive, which means that we do power electronics and compute for ICE, for hybrid or for EV. So it doesn't matter where you invest. All three of them require power electronics and compute. So that helps us in terms of which platform we're on. I would say on the consumer side, I'm pretty much calling it flat. And not sure where exactly it's going, but we're not planning for a lot of growth in this space. We mainly do high-end consumer products. And -- but still, I mean, it's not a space that has high growth, and it's not a big focus for us.

Okay. You mentioned that you're in about 30 countries. How is Flex leveraging its global footprint to support regionalization strategies for customers and to mitigate tariffs?

Yes. So the regionalization thing started 5, 6 years ago, right? It started with the conversation around all this entity list and that started in round one when Huawei was an issue and things like that. It became a bigger thing with the pandemic and where resilience was a big conversation. And then with kind of tariffs 2.0, it's become a bigger conversation. But in the end markets that we're interested in, that journey had already started. Like medical devices is a great example. I think people have been regionalizing for the last 5, 6 years because you wanted to build local supply chains. Industrials were always somewhat regional because they're big and bulky and hard to move around other than in electronics, right, which could have been mainly in Asia. I'd say the place where the conversation accelerated in this round of tariffs was in the consumer end markets. But it's not a place that we put a lot of capital in. It's not one where the math actually makes a lot of sense. And so I'd say people are investigating a lot. I'm not sure there's been real movement in terms of consumer end markets and regionalization. So how we think about regional footprint in general is we want to move our footprint to the highest growth areas with the best return, which means that certain regional footprints have been growing for the last 5 years and certain others are shrinking. And that's the nature of what's happening globally. So for us, North America is our highest growth kind of regional footprint and followed by Europe. And then we keep balancing out all the other regions if it doesn't have growth.

Okay. You've expanded margins consistently over the last several years. Will you continue to see that expansion? And if so, what are those key drivers?

Yes. I'd say even if you look at our most recent guidance, our fiscal year ends in March of next year. So you can kind of read into the second half guidance we have given. We're north of 6% as we'll wrap up this year. We do Investor Day every 2 years. So we did it a couple of years ago. Next one is coming up in May. And that time, I talked about a 6% operating margin for next year. We obviously lapped it a year ahead of schedule. So that's good news. And so we'll have to rechange our guidance in May when we come to Investor Day again, and we will. But how I think about margin is this way. Is if you think about the Flex story on margin, it's all been about mix shift and productivity enhancements. If I look forward, I would say, think about the big growth areas for us. Mix is going to continue to shift, right, big into certain areas like data centers and utility. So that mix shift is going to continue to happen. And that is margin accretive for us. Productivity, which last 5, 6 years has been about better capacity utilization and all of that, and we have had a huge productivity benefit, I would say, accelerates in the next 5, 6 years because we have barely started on the journey of productivity due to AI. And I feel that's a game changer for industrial companies like ours. So if you think about margin moving forward, yes, we've gone from 3% to 6%. But somehow, I feel like our journey has just started in terms of margin acceleration, and we'll give updated guidance in May.

Okay. I love that. I do, with a few minutes left, open it up if anybody has a question in the audience. If not, I have one burning question. Cool. All right. So you touched upon a bit on kind of -- in your last answer, but it does seem like there's a new announcement every day about AI CapEx investments. How and when do you see that benefiting Flex?

Well, it's already benefiting Flex, right? Two years ago, in my Investor Day, I said data center was going to grow 20%. And then right after I said it, first year, we grew 50% on top of big comps, we're growing 35-plus percent this year, right? So it's already benefiting us in a big way. But I'd say in terms of new announcements, the way to think about it is most of them take a couple of years for capacity to get added in and for it to actually ramp up. There may be some early on investments that customers may do for power and things like that, but it takes a while to build these things and to ramp it up. So we're looking at investments now for what will show up as revenue for us and for customers in 2027, 2028. So I look at those infrastructure announcements as things that we're talking and negotiating with customers on how to add capacity to build up for it. But in a lot of ways, it is real because think about just data center U.S., right? A few years ago, we were like, what, 20 gigawatts data center capacity. We're talking about the next few years going up to 100 gigawatts, right? And just that 20 doubled in the last year. And I've been doing this for a long time. I've never seen a year where we doubled. And so going to even something as simple as 100 gigawatts is going to take quite a bit of investment, not just in terms of physical capacity of data centers and the power required for it, but everything downstream. For us to build and test is going to require capacity and investments. For us to do a rack that used to be 10 kilowatts now is going to be 1 megawatt requires resources in terms of engineering, capital equipment to test and deploy. So it's really end-to-end investment that has to go in for these data centers to have a play. So I would say without getting so caught up in this $8 trillion investment number that everybody is talking about, in the next 3 years to get to even 100 gigawatt of data center is going to take quite a bit of investments. And a lot of those discussions are already in the bag in terms of what we are seeing and getting deployed as we speak. So I feel pretty good about it.

Sure, very much so. I guess is there anything we haven't discussed in this conversation?

No, I think we've covered a lot. I think the important takeaway, I would say, would be, first is if you think about Flex as a company today, still a big contract manufacturing company, deployed pretty significantly into data centers and utilities, fastest-growing sector, but we're still a company that does amazingly well in the contract manufacturing space by moving into higher-end goods and improving our margins. But we're the only company who can really integrate end-to-end in the data center space and do it at scale and deploy it with speed with our own technology that I think is a differentiator.

Well said. Perfect. Well, thank you so much for your time today, and great conversation.

Thank you, guys. Thanks for having me.

Thank you all.