Flex Ltd. ($FLEX)

Thank you joining us. My name is Amy Lu. I'm the TMT specialist at BofA, and I'm stepping in for Ruplu here to host the session. So thank you, everyone, for coming in. And we really want to thank Revathi and Michael for joining us today. This is really a great time to be talking about Flex. And so we definitely want to talk a little bit about the spin. So I'll hop right into it, so we can really save some time at the end for questions.

Revathi, you really create a lot of value for shareholders in the past with the -- the NEXTracker spin-off. And so can you talk a little bit about why now is the right time to be doing the spin? And what will let you do that you could not do prior as a combined company?

Yes, I'd say first is in the last 7 years, if you think about Flex, where we're a pure-play contract manufacturing company, we started building out this infrastructure to put compute power and cooling together well before the ChatGPT moment. And by the time that came around compute got more power intensive, the technology of this thermal architecture became a big deal. And then growth was explosive, right? So as all those things started to come together, we ended up with this company that has more of a product architecture with all the technology shifts going on and then the compute scaling was happening at the same time. So it felt like we were putting all our money into the data center utilities build-out. Capital allocation works where it's a zero-sum game, the best returns get all the money and we run our portfolio that way. But at the same time, we had tremendous opportunities in our health care and our industrial business, and we wanted to invest in those two. So we reached a point where there was enough scale in both businesses to say, it needs different management attention. It needs different capital allocation framework. So let's kind of go spin that out to make sure both businesses can be successful longer term. So that was the thinking behind kind of why now. So you get out of that. Flex, which is going to be a pretty significant sized contract manufacturing company with some great end markets that we want to really double down and invest in. And then out of SpinCo, you get something of an industrial electrical player in one of the largest electrical transformation that is happening today in the space. And so you kind of get best of both worlds. And so the thinking was, let's go do it and why not now? Timing seemed great.

Yes, great. So we want to dig a little bit into both of these businesses. Let's start with the SpinCo. Can you talk a little bit about what the SpinCo provides in terms of competitive differentiation and then the addressable market and how fast that market is growing?

Yes. So what's cool about the SpinCo is we -- sometimes people are confused that we put everything that happens in an AI infrastructure build-out into SpinCo. That's not what we did. We have been thinking about SpinCo in terms of thermal architecture. So the technology is all about what's going to drive heat, how do you cool that and how do you bring the best technology to put that together? And then take that all the way out from data centers to the grid because the largest electrical transformation, driven by data centers is also going to happen in the grid. So that was the framework and the thinking from a technology perspective. So SpinCo has our cloud business, our cooling business. Our cloud business being rack integration and then cooling business and then our power business, both embedded power and distributed power. So that's what kind of SpinCo has. We have said $6.7 billion last year. It is going to grow at 70% at the midpoint this year and then at 80% next year. So that's what's in that kind of SpinCo business. And it's really driven by the AI infrastructure investment, and that's going to drive electrical technology transformation that will move forward into the grid itself. So that's what's there in that business.

Yes. And so you mentioned you guys guided really strong 70% this fiscal year, 80% in the next fiscal year. Can we talk a little bit about fiscal year '29? What is your visibility there and a lot of investors are really focused on whether there will be a AI kind of slowdown in demand. And so what is the visibility you have there?

Yes. So what I'll start with is to say, we've said '27 is 90% booked out, '28 is 70% booked out because we have flow-through business also that kind of flows through in that 70% booked out. Haven't given '29 guidance, we'll have an Investor Day here in fall where both businesses will give more longer-term guidance than what we have provided so far, but I'd tell you, if you think about kind of CapEx cycles and investment today, there's a lot of leg work that needs to happen to make those investments come true. So we're power-constrained. So when you're thinking about '29, those investments are getting made, thinking about how long it takes to get power 12 to 18 months. So I'm not going to give you new numbers for '29 because I've already given you some fantastic numbers for the next 2 years. But the way you should think about SpinCo in the 3- to 5-year cycle is what SpinCo will come out looking like in the next 3 to 5 years is right now, heavy data center business, AI infrastructure build-out, strong utility business in terms of grid build-out, but really, I would say, think of it in the mid-cycle, long cycle play right now. We will continue to invest as we have done in the last 5 years in building this out into the services business, which we don't have enough of a portfolio today. We'll invest in more kind of product flow-through business that will see an impact from the 800 volt DC solid-state transformation. So SpinCo coming out of this in the next 3 to 5 years, hopefully, will look more like a balanced portfolio in terms of short cycle, mid-cycle, long cycle, no cycle business and more electrical kind of play more than anything else. So that's the best I can give you at this point in time, but kind of hold on for our Investor Day for more.

Great. And can you talk a little bit about the relative growth rates of the Critical Power segment versus the embedded power as well as the margin profile for each -- and what could drive margin expansion from here? Is it driven sort of by the higher voltage abilities? Or what can drive margin expansion?

So what we have said about growth is in the 70% and 80% growth that we have shared, they have said both cloud and power both kind of grow in that same direction. This year, power will grow more cloud will go less. Next year, it will switch, but there -- it's not a barbell. It's pretty close in that range. It's not -- 1 is 10 and another is 150. They're both in that kind of 70/80 range, both power and cloud. We haven't split out embedded and distributed power. But all I'll say is that both pretty high growth. One, driven by this power density technology shift that's happening in embedded power, right? So you're going from 10-kilowatt, 35-kilowatt to 1-megawatt rack to 400-volt DC, 800-volt DC. And our distributed power business is basically driven by we're small, we're nimble, we know how to customize and put things together. So that's having a pretty high growth trajectory. What I'll say about margins is this. What we have said is when we exited last year at $6.7 billion, we were 9.2% operating margin. Cloud is a little less than that. Power is in the mid-teens. We invested around 100 basis points into the business in the last year. We'll recoup that this year. And then next year, we'll grow another 50 to 100 basis points in terms of margin driven by both mix and incremental basically. So I would just think about that framework to say that as power density and complexity improves, it should provide us continued opportunity to improve margins longer term. Power being in the midterm teens is lower than our peers, but we don't have a 100-year heritage, right, for electrical. We just build this the last 5 years. We're putting in a lot of investment to drive the 60%, 70% growth and we expect that to catch up with our peers at some point in time. So margins will continue to improve.

Yes, you caught me on me my next question. We want to talk a little bit about the competitors in the power space. And to your point, the mid-teens operating margin, Delta and Vertiv were at 20% to 25% operating margins. Is that something you could expect to get to over time? And how we -- what are the key levers to get you there?

Yes, absolutely. I've said my history is coming from running one of the largest electrical players. I've seen them go from low 10s to tip them to the high teens and then all electrical peers have progressed from there. So it's not a pricing issue because our pricing is fine. It is about we're building this from scratch, right? We've put together 4 acquisitions to build this electrical business. So we have to make investments. Electrical, the history of that business, in general, it's never grown 60%, 70%. It's all 5, 7 was a great year for us, a decade ago. So it requires heavy investment to build the right infrastructure out. We're trying to build that out in an AI-native way. So yes, it is going to get to the same that our peers is at, but we will -- we look at this in terms of long term, right? We will build out the profile in a way that we're making big inroads into margin, but we're also investing for the long term, because in 2 years, we're going to be one of the biggest electrical players here in North America. So that's a pretty heavy responsibility, and we want to make sure that it's done well. So it will be a mix of good margin improvement, but good investment to build out the base case for the business.

Great. And now I want to touch a little on the cloud side as well. You guys have mentioned Google and AWS announcements in the past. What is your target set of customers here? And what are the key competitive advantages that Flex has in the cloud space?

So let me just step back and say for the SpinCo, the way I think about customers is hyperscalers, neoclouds, colos, silicon providers and utilities. All 5 of those end markets we participate in. In all those 5 end markets, we will give you just cutting metal, if you want. We'll put that metal together and integrate it into trays and racks and do all the way to L11 testing, and we'll drop it in place if you want. We will build your power modules for your chip or will design your power custom rack solution for your chip or we'll do all your distributed power bus bar, data bar all the way to utilities. So we'll provide this entire portfolio for each of these end markets that I talked about. Hyperscalers will be concentrated because 4 of them make up 75% of the current spend. So there will be some concentration towards that, for all of them, right? So -- but the way I think about diversity of portfolio is within hyperscalers, are we providing this wide variety of products? The differentiation for us in the entire space is this you're either an electrical player that you've got the legacy -- 100-year legacy and you have built it in a certain way, right? Or you're only playing in the compute integration space. Today, all those walls are getting divided. When you design your compute integration, you'd want to think about integrated cooling, you want to think about how much power and how do you dissipate that power. So bringing that thermal architecture technology thinking to the customers well ahead of the cycle is a true differentiation, right? So electrical players are trying to enter into that space now, harder to do because margins are the other direction, for us it is easier to do. And if you look at our competitors, we have a competitor, maybe a couple of competitors in cloud integration. We have one major competitor in embedded power and we have a couple of major electrical players in distributed power. We're the only ones who are doing it end-to-end. So if you're a hyperscaler or a neocloud who needs everything designed, we're kind of your only pure-play solution.

And I have to ask what else in the room, AI CapEx spend. What are your -- how do you get confidence that the demand will continue for compute spend? And what are your concerns that it could potentially flow?

I'd say the same confidence that all of you are hearing, right? Outside of the $1 trillion investment, we're hearing this $2 trillion of backlog that these hyperscalers have, right? And so we are the beneficiaries of trying to fulfill that backlog. So there is a moment in time of how much compute requirement is out there and how behind we are to fulfill that. So that's why we have a lot of credibility in terms of the backlog. We -- for -- what we are building out today fulfills what people need in 2 years and 3 years, right? So there is a long cycle in terms of how backlogged the industry is. And for us, the great news is we can use that investment cycle then to take the cash we generate and invest it in the electrical side where we're building out the solid state and 800-volt DC infrastructure. So I would say, I don't have a crystal ball to 5 years and 10 years, but I can tell you this, the next 3 years in terms of compute infrastructure seems like a solid spend. After that, the grid has to transform. Our grid is so behind that's going to be a 10-year play plus for that grid transformation to work out, right? So I'm thinking of it in terms of both cycles.

Great. And so does the SpinCo have enough of a footprint to support these upcoming projects? Or will you need to add facilities from here? And how do we think about kind of the annual CapEx for SpinCo?

So we have announced a big CapEx this year, mainly towards SpinCo. And we have said, as you think about us exiting this cycle, think of it as kind of 3% of revenue. But remember, revenue is growing pretty fast, right? So it will need continued CapEx. I would say most of our investments are here in North America and U.S. And it's more to do with getting power to our facilities, getting cooling infrastructure set up and things like that. So we are actively investing in extra capacity. We've already leased a new facility in Texas, in Georgetown, that will give us 50-megawatt of power. We just announced an acquisition in Iowa for our utility business. We'll be expanding that pretty significantly. So yes, there will be extra investments. But we think we can hold it within that framework of 3% of CapEx.

So in March, you announced that Flex is building the 800 VDC for NVIDIA. Is flex manufacturing or providing everything in that power rack? Or how should investors think about the allocation between you and potentially other peers like Delta?

I'll tell you this that in the 400-volt DC, 800-volt DC space and embedded power, there's basically 2 players, right, us and our largest competitor. We always are 1 or 2 in most of these design cycles, and then we both share manufacturing typically so that they can have some supply chain resiliency. So I'd say, yes, we are pretty not just kind of NVIDIA. I'd say almost all hyperscaler who is designing a 400-volt or 800-volt DC, we are present in that design phase. So I'll leave it at that.

And kind of moving over to AMD. You also announced that you're building the compute tray for the AMD Helios rack. How do we think about the AMD revenue opportunity over kind of the next few years?

I'd say, whether it's hyperscalers or neoclouds or silicon providers. Like I said, we want to be diversified everywhere. So we're pretty excited about the AMD opportunity. We're kind of the North American player for integrating a very complex product for their GPU product line that's going to grow pretty well. So I think it's part of our diversification strategy of how we're playing out this business.

Great. Kind of taking a step back, do you see SpinCo growing more via organic growth or acquisitions at this stage? Over the years, there's been multiple acquisitions. So do we expect kind of your history of M&As continue?

I'd say, yes, you should expect those. It is going to grow organically, 70%, 80% we've already said. Will we use the cash to look for smart technology investments? I would say absolutely.

Great. So we talked a lot about SpinCo. Let's move more towards the RemainCo. You've decided -- you've guided to low single to mid-single-digit revenue growth. What advice are you giving to Michael? We'll get to Michael in a second. But what advice are you giving to Michael, who will be the CEO of RemainCo after the split?

First is I'd say, we are fortunate to be in this position where we are not only able to stand up 2 companies but then have the candidates to succeed both leadership teams, right? That is awesome. And that didn't happen by chance. Michael and I have been side-by-side on this journey for 7 years. And so our -- my advice to him is, of course, I'm biased is it's a good playbook. Hopefully, you continue it. I'm going to hand it over to Michael, so he can talk about the playbook.

Right after you said don't screw it up, I think. So first, I think you had a question earlier also, and maybe I'll just provide some context for what Flex post spend looks like. So keep in mind, even post spin, Flex is going to be a $22 billion manufacturing and services platform. Still operating at a global scale, still servicing a wide variety of diversified end markets. So a very substantial company, very difficult to replicate by others. If you think about the playbook that Revathi mentioned earlier, it's a playbook that look for the past 7 years have led us to this moment, where you've seen top line growth, you've seen us double margins. You've seen the second value creation opportunity, first with NEXTracker, now with SpinCo in the last, what, 3 years, and that's a playbook that we'll be executing into the future as well. And so it's a playbook that really isn't about chasing revenue. It's about generating high-quality earnings that maximize cash and is being underpinned by this constant desire to continue to expand margins. And what that does for us is that enables us to really deploy the capital to the highest value opportunities. And for us, in the portfolio today, we have terrific opportunities that we just haven't been able to get to because that capital has been deployed where it should to the highest value opportunities, and that's been data center up to this point in time. Now post spin, that will be destinations like industrial and health care, among other things. So really excited about where we can take the business. When you think about where the high-value market opportunities are, I would think about those markets that are tied to longer-term secular trends, right? These aren't sort of event based. So our health care business, very substantial medical devices business and drug delivery business, type of things like aging population, increasing chronic disease. Think about our industrial business, lots of discussion around the geopolitical situation, creating regionalization right for our applications in robotics, warehouse automation. Communications, satellite communications for us is a fast-growing market. People's desire for ongoing always reliable communications is another opportunity for us. And then don't forget, we're not spinning all of our data center business into SpinCo. We saw 3 large businesses that are still positively influenced by pull-through demand from the data center. All of our networking business, anything involved with secure communications that could be high-speed switching, it could be optical, it could be interface technologies. That entire suite of networking services is staying. Energy infrastructure. So we're spinning our IP and our power product portfolio, but we're maintaining a contract manufacturing focus on power generation, transmission, distribution and storage. So a large and growing business. And then finally, semiconductors are required to fuel the growth of AI compute and that takes capital equipment. So our capital equipment business is also positively influenced by data center trends as well.

And can you talk a little bit -- I mean you touched on most of the verticals here. Can you talk about where we could be seeing more end market recovery cyclically, and kind of what's most compelling to you right now?

Yes, I would say this. When you think about the portfolio and now these diversified end markets, you have to think about it on spectrum. And on the one end of the spectrum, we have our high-value markets, industrial and health care. And those markets have been and will continue to grow, no recovery necessary. Those are just up under the right verticals for us. On the other end of the spectrum, we've long talked about really deemphasizing our exposure to some of the lower-value markets, things like consumer. And if you remember the story over the past 4 or 5 years, we've taken out over $2 billion of consumer business over the past few years and replaced it with higher value data center business over the same time. And then in the middle, you have things like automotive, right? So automotive, no secret selling into a really challenging environment. What I would call that as a stabilization period. So not going backwards is actually going forward when it comes to that industry. So that's what you get when you have a diversified portfolio, high-value markets will continue to lean in on, and we'll continue to diversify out of some of the lower value opportunities.

And how are you both thinking about the capital structure for each company, the free cash flow profile and kind of uses of cash from here?

Yes. So what we have said is Flex will retain up to 19.9% of equity in spin and then post spin, it will be an attractive way for Flex to use that equity to pay down debt. So Flex will come out with a clean balance sheet, investment-grade rating, SpinCo will be not highly leveraged. So both will have very clean balance sheets. And then if you think about capital allocation moving forward, SpinCo, as I've said, is going to focus on organic growth and M&A. That's going to be an important part of the strategy in terms of capital allocation and Flex our view is we'll continue the capital allocation that's worked for us so far.

Okay. And post spin, are there any dissynergies that investors should be aware of or focused on?

No. We've been running our 6-business unit structure stand-alone for a long time. Our business unit leaders are incentivized to grow and manage those businesses end to end, which means that our factories have also run that way. So out of our 80 factories, 5 of them have some level of overlap, so that's all we have to work to separate. So that work is already in place. So it's not a pretty big deal. And then SpinCo will have stand-alone company cost and then Flex has kind of some costs left over that we know how to manage. So we don't expect any significant dissynergies in our modeling moving forward.

Great. And just my final question. What do you guys think on both sides of the business is still the most underappreciated sort of by the market and by investors?

Yes. I'd say that's one of the more exciting parts is that for sure, there's a lot of excitement, rightly so created by our decision to spend, right? That's a business that is front, center and gets the headlines. But I think what's most underappreciated is the -- really the position that Flex post-spin is in to grow in these higher-value markets and continue the trajectory that we started 5, 6, 7 years ago. So by executing this playbook, it could lead us to potentially another event, another value creation opportunity like NEXTracker, like SpinCo, but for sure, it provides us with the road map to continue to drive top line, but more importantly, earnings expansion over and over again in these higher-value markets that we've leaned into up to this point in time.

I'd say the SpinCo will see what value the market assigns to us post spin. So at this point, we think it isn't valued enough, but that's always my biased view. So I think the market's understanding the significance of creating such a large electrical player. It hasn't happened in a while, right? The stand-alone something from scratch and create something like that. So -- and then I think the largest transformation that's happening in the electrical industry is in front of us and we get to create it from scratch without any of the heritage and history, right? So I'm sure that will have tremendous value too. So yes, we'll see where it goes.

Great. Thank you. I really appreciate both of your time. Thanks, everyone, for joining us.

Excellent. Thanks, everyone. Bye-bye.