Vertiv Holdings Co (VRT) Earnings Call Transcript & Summary

All right. Moving right along. Maybe this one is going to be a little bit more cordial. We're here with -- I'm not going to give you context around that one, with Giordano Albertazzi as well as David Fallon. Gio is the CEO of Vertiv and David is the CFO. I don't care what anybody says. I still think that is one of the most eye-popping, ferocious investment cycles we've ever seen. Nothing has slowed down in the last couple of months from what we can tell when it comes to data center development and pipelines and funnel. And so I'm still extremely bullish on this stock. But with that, Gio, maybe just give a bit of an intro as to how you see things in maybe longer term, and then we'll jump right into the questions.

No. Absolutely, right. I thought you already did a very good intro. So thank you very much, kidding aside...

Well, you agree with it, I think...

I kind of tend to agree. Yes, absolutely. So we are here with investors about what is it, 3 months, 4 months after our Investor Day in November in Atlanta. Of course, earnings call a month ago. We continue to see a very strong landscape for the data center for Vertiv. In February, we confirmed the plan and confirmed the long-term vision for the business and for the industry that we shared with you in November, and we continue to do so. So we are quite positive about the industry, we're quite positive about the pipelines that we've seen unfolding. And again, we go back to a very strong Q4 where we again delivered a very strong quarter. Our ability to deliver on our promise is certainly once again confirmed, and we feel good about our portfolio. The technology evolution that we shared with you in November, some of you saw our portfolio at SC24, things are going in the right direction. And pretty much a story that we shared with you is all the more reasons are strong and well. So with that, I think I would -- let's move on, too.

I know you guys aren't going to guide orders, so I'm not going to ask that question.

Yes, we're not.

But I guess, can you give us -- just remind us of how the lead times on all this stuff work? How -- when you get an order today, when that shipped, how much these customers are ordering when it comes to the -- what is now an average 200-megawatt or whatever it is in the funnel? And to be clear, 200 megawatts compares to probably an average of, I don't know, 30 to 40 probably historically. So these are huge numbers. These aren't average-sized facilities by any measure when you're looking back historically. Maybe just talk about how that all works? How customers order for you and what that lead time is?

Sure. So as you were alluding, clearly, very, very big orders, and some of these big orders have a lumpy nature. They can happen one side or the other side of the quarter. So that's why we say, "Hey, do not focus too much on a quarter number. Look at the trailing 12." And we're talking about 30% year-on-year trailing 12 for 2024. That generated a 1.2 book-to-bill for the year with more or less 30% more backlog going into 2025 than we had going into 2024. So strength. But market is becoming bigger in absolute terms, but also bigger in terms of the individual jobs that are out there. And typically, these orders and the big data centers' orders are done in chunks. So if you think about a lot of the market talking about kind of a 1 gigawatt or multiple hundred of megawatt data centers, typically, orders are cut in chunks. And we receive orders pretty much when the building plan is there when -- very often when the -- actually construction is about to start. So that's pretty much the cycle. In the last call -- in the last couple of calls, we talked about lead time -- a requested lead time for our customers around 12, 18 months. We are happy that our lead times, if customers ask, can be shorter than that and we see that as an opportunity to gain some more business. But that's the way the market is behaving. So larger projects, a little bit lumpier to call and pretty much all the time connected to an actual site that has been built out there.

So how much of the '24 orders roughly? Like how much of that kind of like is filtering into '26? And I assume that it's -- there's some in '27, but it's probably a minimal amount?

Yes. We are not disclosing that specifically. We certainly have backlog for the next year. And as we speak, the backlog for the next year is further growing because of the lead times, the lead times that our customers request pretty much or most frequently. So '26, of course, was there already at the end of last year and it's building up. So we feel pretty good about how backlog is unfolding for the year and pretty much reinforcing the model that we shared with all of you.

And I think you had said you were bullish on pipeline and in reference to '26, '27, how do you -- just remind us of how you define pipeline and how much -- how visible that business is?

Sure. So pipeline for us, or funnel otherwise called, is all the opportunities on which we are working, the opportunities that are real, that have a commercial. Most of the times, a quote attached to that. So it's something that we look at religiously. I look at that weekly, monthly, very religiously. And so it's across the entire organization. The direction of travel that they indicate is certainly supporting our plans, and it's very encouraging. So it continues to be strong in that respect.

Can you just step back and just try to put some of this news flow into perspective? You guys talked a bit about some Europe pushouts from an orders perspective in the fourth quarter. One of your peers came out and talked a little bit about that as well. Maybe just address that from a backward-looking perspective first. And have you seen anything there that changes your view on what the direction of travel is there in Europe?

Pipelines for Europe are strong. So we are happy with what we see. We see strength. But there could be, as we experienced in Q4 last year, some delays. There were some delays that we experienced. Certainly, the market is a little bit more, how can I say, regulated a little bit with more red tape than, for example, in North American or Asian or the Middle East. And that is a drag to the speed at which the market can develop. But having said that, that's not new for Europe. We see a lot of good signs in terms of -- also the general awareness, if we think about the Micron initiative, the conversations that are happening at European Union level to unleash the potential of AI. So 2 things. The environment seem to be getting more, let's say, friendly and conducive to AI and to an acceleration in AI infrastructure, and that is corroborated by today a strong pipeline. So we are optimistic about that. Let's see how it unfolds.

And then just to keep it -- to kind of wrap the global discussion. China, obviously, a lot coming out of there when it comes to DeepSeek, but then Alibaba and their $53 billion or whatever that is CapEx number, it seems like things -- they kind of press the button midyear last year and things are really accelerating there. Nobody really talks about it, but maybe just your update on China and in Asia...

Yes, exactly. It's never China alone. Already about 6 months ago, we started to talk about China as kind of green shoots and positive direction of the data center, at least the data center market, and we see that happening. The news that you were alluding to, Steve, are exactly going in the same direction. So certainly, it continues to be a competitive market as the nature of the market in China, but certainly positive in terms of market trend. And not only that, but also the very same kind of, let's say, enthusiasm of the Chinese hyperscalers reverberates in the rest of Asia with investment in infrastructure in the rest of Asia. And it's part of the optimism that we have for Asia in general.

And you're pretty confident in your market share there that you can sustain your market share in China. You talked about it as a competitive market, but you guys have been there for a long time and you're pretty strong there?

Yes. We've been there forever. We are a Chinese company in China with a Chinese supply chain with Chinese technology. It's not only a place that is extremely important for us to compete in the Chinese market, but it's also very important for us to compete and be very relevant in the rest of Asia. So I think we have a good position there. And playing in what is probably the most competitive market in the world is very important because it keeps you extraordinarily fit and ready to compete across the world.

So these are all public comments, obviously. So it's not like you're talking about a customer that something they haven't said publicly, but Microsoft came out on their conference call and talked about pivoting their CapEx a bit to the shorter-lived assets. There was some news about them moving some leases around 200 megawatts in the grand scheme of a 100 gigawatt type of market is, I think, kind of silly, but how should we interpret these data points, because they're obviously not positive, and there's obviously an abundance of positive data points out there to offset them? But like how should we be interpreting when these -- when your customers say this kind of stuff? Help put that in perspective and what it means to the market?

Yes. I don't believe that, that changes the outlook for the market as a whole. The market is very dynamic. And there could be different roles that the different players want to play over the next few years. And there are a lot of actors. Clearly, the large hyperscalers are -- have a very important role. We heard very recently their CapEx declarations that are certainly encouraging for the industry and for us in the first place. So if someone decides to invest more on one end of one type of, let's say, CapEx and less to another, there is more than enough people that step in and offset that. Again, we see that as perimeter changes and less as in or out of the investment cycle.

And is there any concern like, for example, I think it was yesterday or maybe the day before, OpenAI stepped up and committed a $12 billion commitment to CoreWeave, whose main customer is Microsoft. Is there any concern that you're really isolated -- not isolated, but levered to one particular hyperscaler that we should be watching out for? Seems like in all your materials that you're pretty broad-based across all these guys. Is there anybody that we should think of that guy cuts, that's really not a good thing for Vertiv specifically? Or do you, in general, pick up your market share at the other end if somebody is actually building the same data center?

We have a very, very diverse customer base. And that's true for hyperscalers, for colocators. So if you think about the entire world footprint, but even the North American footprint, it is a very large array of players. So I think we see these dynamics as an opportunity to, if anything, have more vibrancy in the market that is already vibrant, but we are pretty well spread across.

Going back to the orders numbers a bit, and I think you guys said this at a competitor conference recently that book-to-bill would be above 1 in 2025. I mean that could imply if it's just above 1 that your orders are actually down year-over-year or at least like even just using the run rate of 4Q, you should be at a comfortably book-to-bill above 1. Any sort of other framework you guys would like to give on the orders?

That would mean guiding orders, which is something we do not do. But again, I think I'll reiterate the message, book-to-bill bigger than 1. I strongly believe that the outlook that we have shared is very alive and very, very well. And yes, that's really the story and pipelines that continue to be stronger and showing signs that again confirm the model that we shared with all our investors.

When it comes to the seasonality of these orders and the timing of these orders, how is this different than the past? Is there seasonality to the orders? Is there a rhyme or reason to why a 200-megawatt facility gets booked in March versus April? I mean, typically, construction when it's kind of flattish has its normal seasonality. So how is this different? And how are you guys gauging?

From an order standpoint, I think that the days of -- the old days where there was clearly kind of a seasonality that repeats over the years and the years and the years is now there. It's more -- the industry is having its own dynamics, and the dynamics is probably very much specific to the individual customers. So as we said, orders can be lumpy and that is pretty much the rule. We like to talk in terms of trailing 12 months because that is more the way represents better the way the business work, the long lead times that characterize a large data center business, but I wouldn't call it anything specific seasonality. Like it's always quarter X that has some characteristics that change over time.

Right, which obviously makes the year-over-year comparisons somewhat irrelevant? I think what matters in any given quarter?

What in any given quarter to me, it is pretty much true.

Yes, okay.

That's why we talk longer than a single quarter, we talk trailing 12.

Right. I think that's what's interesting about this is you guys are booking $2.4 billion in orders. And if you do the math around what one average 200-megawatt facility means to that order book, $3 million per megawatt is -- can be massive quarter-to-quarter. People don't appreciate the scale. I think I blame you guys were putting out like hunting us for putting out hundreds of gigawatts. It just sounds like a number, but 1 gigawatt is a -- that's $3 billion worth of TAM in a $33 billion TAM market.

Yes, sir. Exactly.

So just from a competitive perspective, when you get designed into or when you're part of a rack for these customers, they're designing you into the rack, how fluid is that? So for example, Dell had your in-rack CDU. They put it on LinkedIn, and they were highlighting their rack and you could see the little Vertiv sign there. Is that for the time being like your socket? Or can they easily just kind of like it will all depend, they can just pull it out and put another one in and you'll get your fair share of whatever they sell, but who makes that determination? Is that Dell? Is that the guy who's building the facility? The owner? So what firm is that socket, I guess, is the question.

First of all, I would probably recalibrate a little bit the thing. In terms of when we talk about a rack, CDU as your example, in a rack, then we talk about a relatively small part of the market and the portfolio for 2 reasons. One is there is so much more in that 3 -- actually $2.7 million, $3.5 million ton per megawatt, then that's little shoebox, if you will, inside is very important, very critical inside the rack because there is the entire powertrain, the entire thermal chain, the chillers, the UPS, switch gears, blah, blah, blah, all that stuff. But even when we look at the CDU inside the rack, and to your question, it is -- the determination is a mix, as a matter of fact, is a mix between the end user and the rack integrator in this case. So yes, I have to influence, of course, both. While we see more and more happening is that the very CDU focus, zooming from the large Vertiv portfolio to the liquid cooling, which is an important part that we like a lot, but it's still a relatively small portion of the total. You zoom in and that rack-mounted cooling distribution unit is becoming a row-based cooling distribution unit, CDU, because, again, more and more, the space within the rack is utilized for servers. So you see that liquid cooling is becoming less of an IT, let's say, scope delivery and it's becoming more what it natural is part of the mechanical infrastructure. So either way, we are very strong. And certainly, the mechanical part of a delivery of an infrastructure is what we have done forever extremely well. So there is -- there are dynamics around that beyond the choice that a certain system integrator can make.

But obviously, they have to all be on the same page when you pitch them. And is it likely that, that customer will choose different for their facility, they'll choose a bunch of different suppliers for that just to kind of like leverage you guys against each other? Or are we not...

Well, it really -- that's really specific to the policy that a customer may have. There are customers that want to have Vertiv only regardless if it's a kind of heat rejection unit or UPS or a rack mount CDU -- sorry, yes, CDU or -- and there are others that will have multi-vendor type of strategies.

Where are we in this technology cycle? Are we -- I think everybody has been -- the term normalization is, obviously, during COVID, lots of backlogs. We've normalized from that. But where are we in the context of normal for this data center build-out? And I think about kind of the cloud cycle where you start here and then everything generally goes to just a steady state and it's about efficiency and productivity. Where are we from a technology and build-out perspective?

Both cases, very early stage. Both cases, very early stage. If you really think and think back, we started to intensely talk about AI and AI infrastructure 2 years ago. That means that given the cycle of data center build, it's probably -- they've started to be built a year ago, hardly, and probably are coming available now or been coming available in the last 6 months or so. So it is very early stage. Not only is early stage in terms of the build-out for AI, but it's also early stage for this technology. There is a lot of headroom for silicon evolution. So the GPUs today and the GPUs 5 years from now will be different. And performance of the GPU 5 years from now will be incredibly higher than they are today. And today, they are incredibly higher than they were 2 years ago, et cetera. But also the technology that goes around it is changing. Again, let's really think in terms of -- I think nothing explains the story better than the density per rack. Historically, in your very relevant cloud analogy, we're talking about loads per rack around 10, 15 kilowatt. We are already -- today, we see a lot -- the majority of the design between 150 kilowatts and 300 kilowatt per rack. And we know that there is a road map out there to go past 0.5 megawatt per rack. So that is a lot of density coming. And with that density comes a lot of efficiency and productivity of the GPU and the IT itself. So that's the primary reason why this race is not going to stop. And with that, the -- let's say, the infrastructure that is not just the cooling, but it's the cooling entirely, the entire power thermal chain, so including the air part, including the heat rejection part, the chillers and whatnot, but also the power are morphing around a much more challenging design. And that challenging design means a lot more value opportunity for the likes of Vertiv. And that's the landscape that we love.

Has there been any change over the last 6, 9 months as to how much visibility you have on these upcoming rack the upcoming rack ecosystem evolution? Have you -- you were kind of early on getting a look at the liquid cooling and what's happening with Blackwell. Clearly, a key partner to NVIDIA. Has anything changed on that front where, hey, they're going to a different level now, so they're like we don't really need you guys anymore. We're going to kind of use something else. I mean have you seen any change?

No, I would say that the partnership with the key silicon vendors certainly with NVIDIA, but also the partnership with our big colos and hyperscalers, all the more reason important now than it was before and the intensity of collaboration is bigger. So just like for future silicon, we have -- are working together, make sure that the infrastructure is ready from a technology standpoint by the same token, working with many of the hyperscalers in their labs and defining what their infrastructure will look like 3, 5 years from now from a cooling and electrical infrastructure to manage the new loads that are coming.

So the market share -- you're confident in your market share...

2 I'm pretty confident. And as we were saying, for us, it's growing markets and a Vertiv that grows faster than the market.

Can you price for this? Historically, when cloud matured, they obviously took all the technology and they started working you guys a bit on price. You're implying that we're still very early in this process. Therefore, it's not really worth their while to get an extra couple of points out of you guys. They care more about you being there for them and allowing this technology transition to happen. How does this kind of play into the price discussion longer term?

So let's make no mistake. Clearly, our customers are very kind of savvy commercial operators. So I don't want to sent any different message. And it's natural they are because very often, they are very large and very well equipped with talent, let's say this way, and experience. But also true that the ability to price really is based on the value that you create for a customer. So it's reliability, certainly is a service level, but first and foremost, long term is the technology that you make available. So you can only price the value -- the extra value that you create for a customer. So that's why we're spending an increasing amount of money in R&D. That's why we are embedding ourselves from a technology development standpoint with our customers. And we understand the industry. We understand the industry very well. So we understand how to create value. And value could be an enablement of a new technology for them. The value could be making the actual building a data center a faster and easier and more cost-effective process, all elements that add to our pricing power, but we like our pricing power when it delivers extra TCO or extra value for our customer, first and foremost.

And you haven't seen anything recently in this kind of delicate dance with your customers that would make you concerned at all about the margin targets you guys have put out there, which implies 30% to 35% incremental, something in that range?

What we see in the markets, the things that we're doing, the things -- the technology, the direction of technology corroborates the plan that we shared with all of you.

Any questions out there? Yes. You can just shout it out, I'll repeat it. Question is with the stock down, why not do an accelerated buyback?

Yes, I'll take a shot at this. So the last 2 years, we've shared our philosophy as it relates to capital deployment. And by the way, we're in a much different position today than we were 3 years ago with our ability to deploy capital, and that's absolutely driven by the improvement we're seeing with free cash flow, which is a separate story, and I would love to get a question on that. But from a capital deployment perspective, we -- the 3 primary levers are dividends and we have a $0.15 per share dividend. We're not going to be ultra aggressive with that at this point. As you have M&A. And we like where we are today from a cash position and it's always important to keep your powder a little bit dry there. And from a share repurchase, what we've shared is that we'll continue to be opportunistic. So we have a $3 billion authorization. We got that in November of '23. The first time the window opened, we did a $600 million share repurchase at a pretty good price. So we will not be timid in doing share repurchases. It is not just something sitting on a shelf, but we'll take into consideration everything else that's going on in the environment. So probably not a robust answer or a definitive answer. We are going to do X or Y., but we'll continue to remain flexible and opportunistic.

The question is IRR is much more attractive than M&A given where the stock is, in your minds?

Yes, that would be a factor we would consider. But at this point, we would just reiterate what our philosophy is with capital deployment.

I just want to very quickly -- sorry, just bang out the tariff question, and we can go a little bit over. We can go a little bit over. It's fine. Can you just give us the most up-to-date view on not what's going to happen with tariffs, but how you look at what has been announced? And maybe provide some context on how differently you're approaching this situation versus what happened a couple of years ago during COVID, because I still think there's some institutional memory in the market about what happened back then with price cost?

Yes. Of course, we will not try to predict what will happen with tariffs. That will be a very daunting challenge. But I think it's important to say that in the last 2 and more years, we've been acutely focused on making our supply chain much more resilient, and of course, also more regional to serve a region for a region. And that is a good place from which to look at the dynamics that are happening in a market from a tariff standpoint right now. Clearly, we are in a different position also when it comes to our ability to manage price, and we have demonstrated that in the last couple of years with quite some meaningful -- there is a very meaningful progress in that respect, progress in terms of our ability to value price our products back to the comments that I made a few minutes ago. But also rigor and discipline in the process that is unprecedented. In general, our ability to manage price in a more favorable way, but certainly in a very, very coordinated and cooperative fashion with our customers has improved. And again, this is a much better starting point than we had in the past.

And are you guys -- just to check the box, are you guys covered by USMCA, because I think Mexico is kind of the area where you do have some capacity down there?

We do have capacity in Mexico, of course. You guys have seen the map of where our plants are. So that's clearly. And we've been operating in Mexico for more than 2 decades. So we certainly have a lot of experience. We know we know USMCA very well. So I think we're in a good shape.

Okay. Then one last question here. Question is, is there any aspiration for an investment-grade credit rating?

Yes, clearly. And we would say we probably have the debt metrics and the balance sheet to actually be IG today. So we got to upgrade it twice last year, and we're 1 level below where we need to be from an S&P perspective, and we'll continue to manage the balance sheet with the aspirations of being investment grade, yes.

I think you guys should lever up and buy back stock, personally.

That's the other side of this. But we do have the flexibility based on not only the cash flow that we have on our balance sheet, but the expectations to continue to have strong free cash flow to be able to do both, right, to successfully deploy capital, while keeping a fairly conservative leverage ratio, so -- which is definitely, we think, qualification to be investment grade.

Great. We're out of time. Thanks a lot, guys.

Well, thanks a lot.