Should we go ahead and get started? Great. Okay. Hi, everyone. I'm Lance Vitanza, senior analyst at TD Cowen, covering telecom equipment. Thank you all for coming. I'm delighted here to be with the senior executives of Vertiv. Let me introduce Giordano Albertazzi. Am I pronouncing that?
Perfect, Lance. Thank you.
Gio actually started at the predecessor company back when it was Emerson Electric in 1999. He was running a plant. He worked his way up. By 2016, he was the President of EMEA. He did such a great job running EMEA, that in 2022, they made him President of Americas, and then obviously, he took over as CEO earlier in the year. Thanks for coming. We're delighted to have you.
Thank you for having us.
Of course, longtime CFO, David Fallon. David, I'm not gonna list out all your accomplishments other than just to say that you were there to help guide the company through the LBO, or through the through its history as a highly levered post-LBO company, and then obviously, its transition to a publicly traded midcap equity. Thank you for being here as well.
Yeah.
I wanted to get into. Maybe just to set the stage, could we talk a little bit about the demand environment? I know obviously we're gonna get to AI. That's obviously on everyone's mind. I was hoping, you know, if we go back before NVIDIA, like two weeks ago, right?
Absolutely.
There was so much angst about hyperscaler CapEx, and the sense was, it was almost that there were investors that were under the impression that we would never see another data center built. You know, that was not what we were hearing. When we talked to the colo, the colo guys, the hyperscalers, the consultants, we were getting the sense, our checks were suggesting that you would see a slower rate of growth, but still positive growth. Again, this was before the NVIDIA news. I'm just wondering, you know, what are you seeing out there, and how would you describe the environment?
Absolutely. If we really want to go before the NVIDIA news, of course, we can very well go back to our earnings call about a month ago. We were describing an environment that continues to be positive for us, and we know we are in a market with a strong secular trend. Data, which is 70% of what we do, has to do with telecom and data center. Data, the underlying demand, will continue to grow, and the demand for infrastructure will continue to grow. Already then, we were talking about some changes, evolutions toward more high density, AI-driven.
We continue to believe in that, in that scenario, clearly, and it's in a scenario that before an AI acceleration sees, let's say, an high single-digit growth of the market, and different parts of the world move at different speed. As I said, we are in a phase of normalization of order intake because lead times are normalizing in the industry. You know, very positive about the underlying demand and very positive about the industries that we serve, and definitely about the data center industry, despite the very big acceleration. There was no such thing, even before the NVIDIA news, of a kind of a massive slowdown or a slowdown to start with.
Mm-hmm.
There continued to be growth in the sector.
Yeah, in fact, just to sort of go a little deeper in terms of, I think you just said, high single digits kind of market growth for the suppliers, right? To get to that, you need what? Does that require mid to high teens growth in underlying data consumption?
That-
-right?
That's a fair assessment, absolutely.
The point being that if we're in a world where data consumption is going to continue to grow, you know, at 10%-20% by definition, the only way that can happen is if we continue to build out capacity at data centers, which is going to require power, cooling, and so on and so forth.
Absolutely.
Okay.
That's certainly the case.
I guess within the overall market growth, are we still seeing the shift from the enterprise on-prem to the sort of the cloud hyperscale, you know? Where are we in that transition? I think, you know, not long ago, it was sort of 70/30 in terms of enterprise versus cloud hyperscale, colo. Has it gone beyond that, or are we sort of 65/35 now, or?
I'll say that that shift will continue but it will be gradual. We do not believe that it will be ever 100%, colo or hyperscale, cloud. We believe that it will stay balanced. Anyway, the shift will be slow and continue to happen. Both markets, or three markets, the hyperscale, the colocation, and the enterprise in very decent health and very well penetrated by us, all three of them.
Presumably, though, there's one thing I think that, you know, we sort of learned when we were doing our work, Vertiv is a beneficiary of the trend toward consolidation, right? In other words, yes, you know, as the demand becomes more concentrated, you might lose some pricing power. On the other hand, there's so much more that is required of a supplier when you're selling-
Absolutely
... globally, is that an accurate assessment, or how do you-
That is a very accurate assessment, but let me elaborate a little bit on that.
Please. Yeah.
Clearly, the colocation and the cloud spaces are global and increasingly global in their very nature. There are, of course, a lot of colocation, kind of, new companies that have popped up in the last 3 to 5 years, but they all expand and become global in their footprint. Our ability to be with them wherever they are in the world, giving them a very local experience in terms of execution, but very coordinated experience in terms of technology, account management, and supply chain, is certainly a great advantage. Going back to the more local nature of enterprise, now, We operate in more than 130 countries, so we are very local and truly global.
Our ability to reach out to local players is second to none, I would say. Very positive about that. We can play the two spaces.
Mm-hmm.
Very, very well. The technology is becoming more complex over time. We'll talk about AI later, but the technology is becoming more complex for everyone. The energy efficiency example instances are more, let's say, important for everyone. The ability of interfacing with a, with the expert in the market, such as we are, is very important, regardless if you're kind of a global player or a local player.
The last thing I wanted to do before we get into AI, is just go a little bit deeper on the three regions in which you operate, EMEA, APAC, and the U.S. or the Americas. I assume everybody in this room has read the transcript from your last earnings call.
Cool.
-without rehashing that, is there anything that you can tell us about each of those three segments or any of those three segments that may have changed since we last heard from Vertiv?
No big changes. I mean, we go back the month, we were there a month ago. The very things that we defined and we said then, you know, we refer back to our guidance a month ago.
Okay.
That's four weeks ago.
Okay. On AI, let's just jump in. You know, our understanding is that a current, like traditional, I'm gonna call it a traditional rack server, is consuming 6 kW to 8 kW of power, something in that ballpark. Our understanding is that in order to facilitate AI, the new racks are going to need to consume 5 x as much power. Is that the right sort of ballpark? I'm not asking you to kind of verify that specifically, but.
Well.
How should we be thinking about the increase in power that is required? I'm sorry, the increase in heat-
Yes
that's generated?
In we call it power density.
Yeah.
The consequences are, heat, of course, and power.
Yeah.
You know, bringing power to the rack. I would say that 6x or 5x or 6 kW is on the, let's say, very prudent side.
Right
-an understated side. We have conversations with some server vendors and chip manufacturers that talk about all the way to 80 kW, 100 kW per rack. It's a totally different world, the one ahead of us. It will be gradual. It will be gradual in terms of penetration in the install base, and it will be gradual in terms of growth. Different chip generations, different GPU generations will have different heat loads, but certainly in 2, 3, 4 years out, we will see definitely a shift towards very high loads. The very high loads require a lot of heat dissipation.
Multiple technologies competing in this moment, the most likely winner is direct-to-chip cooling, that will be either a single-phase or two-phase liquid cooling to chip. The interesting thing is that when you cool directly on the chip, say, an 80 kW rack, but also a 60 kW rack, you have a lot of heat being to dissipate around the server and the rack that is not directly extracted, let's say, from the direct-to-chip. You will have racks that generate heat, that you have to remove through with air, at levels that are higher than the.
Hmm
that we have today.
In addition to-
In addition-
-traditional-
In addition to.
With, with direct to chip
direct-to-chip. It's very interesting for us, of course, for two reasons: One, the whole system become more complex.
Mm-hmm.
We like complexity because, you know, you have to mix technologies and orchestrate through controllers and through management. You have to manage inside the rack and across multiple racks and the system. We like the technologies, we like the complexity, but also it raises the bar as to what companies can really compete in that space.
Yeah.
can compete and scale
Right
In that space. It's a global play, so.
It accelerates that transition from the big three or the big four having half the industry. You know, in AI, the big three, the big four, maybe the big five, you know, they're gonna have 90% of the industry. Is that? I mean.
I don't know about that exactly, but, you know, we expect to, we know we are very well-
Your share is going to go up in AI?
We believe we are very well positioned for the reasons that I explained.
Then in terms of the product, do you feel confident? I mean, is this, you feel good because you know that you're going to wind up getting there, and that by the time it's needed, you will have the right technology? Or do you actually have the technology in place today to serve this opportunity? Or I guess, to what extent...
Sure
... do you have the technology?
Sure. Again, it's not one technology, it's technologies.
Right.
They need to orchestrate the multiple technologies on the same type of.
Sure
...problem. The technologies are immersion cooling, or direct-to-chip. We believe that direct-to-chip will be the winner, we believe because we talk with the server manufacturer and chip manufacturers. We, of course, have a lot of engineering going on. We have high levels of cooperation with many of the vendors in the space. Yeah, we feel confident. We will... We have the technology, and we will have the technology as the chip technology evolves.
Where do you think Vertiv stands relative to, say, Schneider and Eaton, with respect to your abilities and how long you've been working on the technologies that will facilitate AI?
Yeah. I would say we're talking about thermal, so it's not necessarily the same type of competitive landscape that we have in power.
Sure.
Some of the players you mentioned are power. Let me not necessarily draw comparisons, here. I'd say that we feel confident. We are strongly positioned from a technology and an ability to execute standpoint.
Okay. I know you touched on this a second ago, but in terms of the timing, and obviously no one has a crystal ball, but, just in terms of thinking about an endpoint, I've heard some people say that virtually all data centers in the future will need to be able to handle AI. I've heard other people say that, "Well, no, AI, if you think about like, I don't know, 10, 20 years from now, perhaps 20% of data centers will be able to handle AI." You know, what's your own view in terms of where we sort of wind up at some point?
My point of view is that it is more complicated than that.
Okay.
In a sense, it's not necessarily binary. It's not kind of AI, not AI. There is a lot of AI today, a lot of AI at work. What we're talking about is the ChatGPT type of AI. It's yet another level...
Mm
... of AI and another level of infrastructure. We believe that that transition will be gradual, in terms of some percents over time, but where it will exactly be 5, 7, 8 years from now, hard to tell. Certainly, there is a big install base of data center and a lot of capacity out there. The need to upgrade the existing stock certainly will exist, and it's something that we are interested in as well. We see a positive market landscape, all said and done.
And you've given me a really good feel for, I think, the volume and the revenue opportunity, and the market share opportunity. I guess, you know, the other question that we get a lot is, well, what about the costs and therefore the margin in sort of AI and maybe new ChatGPT AI versus, you know, some of the traditional stuff? Is there anything that we should be thinking about there, or is it really similar cost, similar margin profile?
Technologies are still at their infancy.
Mm
... in many respects. Volume is definitely insufficient to be able to draw any conclusions. Again, it is higher technology, more complex orchestration, dare I say, more value addition from a vendor like ourselves. Where that leads us, early to say. You know, we like what we see ahead of us.
Sure. Okay, let's switch gears and talk a little bit about supply chain. Still an important topic. I guess, maybe just at a high level, you know, where would you say we are in terms of the road back to normal? Rather than focusing on price, cost, headwind, tailwind, which we'll get to in a second, but really, is it possible to talk more in terms of, you know, how much product are you buying on spot? You know, is there a, today versus say, you know, at the height of the crisis, you know, the number of products or parts that you're having a hard time finding today versus, you know, 6 months ago, the length of time and the delay today versus 6 months ago, those types of metrics?
Absolutely. I think we were vocal during earnings call about our spot-buy situation, and the spot-buy situation is pretty much normalized. Normalized means we're going to kind of the pre supply, general global supply chain chaos. Now, there clearly are some commodities that are still tricky, as we said during earnings. We're talking about power electronics, some power electronics and some fans, again, a victim, if you will, of the same power electronic shortage, but normalizing. There are two aspects of or maybe more than two aspects of the supply chain situation. As we said a month ago, on the one hand, is the world is normalizing, which is good news.
General supply situation is normalizing. We have done a lot to move ourselves to a much more resilient position than we were before. Like many in our industry, we used to have probably one vendor, very often one vendor per our most critical components. We were saying, you know, critical component, you want to be with the strongest vendor, if you will, or partner.
Mm-hmm.
You know, we now understand that that type of position that was very risky in terms of resilience. We have now almost all the critical components on a dual or triple source, and that's a philosophy that we inject in everything we do now. We're moving this kind of a logic further in a sense that it's not just enough to have two suppliers, but we wanna make sure that there is supplier resiliency, geographic or geopolitical resiliency as well in what we build. It's a new philosophy for us to handle supply chain and to design products.
Supply chain, generally speaking, for us, means also our ability, you know, to manufacture in terms of a net capacity. Like many, like the whole industry, you know, we were behind demand in terms of available capacity. We have and we are continuing to invest. We opened a new factory in Nové Mesto... Sorry, that was my plant, you know. Hence the Freudian slip here. In Monterrey, we opened a new factory in Monterrey, Mexico. That's our largest factory for thermal management-
Mm-hmm.
hence very relevant for everything AI in the future. We expanding capacity in Reynosa, Mexico. We are expanding capacity in the UAE factory in Slovakia. In India, we're building a large thermal factory in India. The other factory expansions that I mentioned are more on the power side of things. We are on a very good trajectory in terms of the capacity we can make available for the industry.
Did you say, I heard you say Slovakia? Did you also say UAE?
UAE.
Okay, gotcha. Okay. It sounds like then you're feeling better positioned. The strategy is not let's just, you know, "Hey, we got through it. Let's just hope that we never have that problem again," this time.
No, no, no. I think everyone learned a lot of, a lot of lessons.
Yeah.
We certainly did learn a lot of lessons. One of the strategic priorities I gave to the organization is operational excellence. It's very cliché, if you will, very simple, but I wanna make sure that the entire organization is focused on execution, and execution for our customers. Again, don't forget, I'm an operations guy by background. For me, that is absolutely important. General statement, but very, very clear actions behind. We wanna make sure that we are, you know, the strongest player from an execution standpoint in the industry.
Gotcha. Last question on supply chain. You know, you have a nice tailwind now from price cost. How long, realistically, do you think that that lasts? I guess, you know, at some point, are you seeing your customers starting to say, "Hey, well, supply chain is easing, so how about discounting me again?" How does that look? How does that work? How do you expect that to look?
Generally speaking, clearly in very inflationary times like the ones that we have experienced, we have all experienced, the price to compensate inflation type of argument is clearly strong. As inflation diminishes or normalizes, well, and by the way, it's not happened yet.
No.
There is still a lot of inflation, and anyway, in any way, but cost of material is not going backward necessarily. As things normalize, of course, the ability to price to offset inflation diminishes.
Okay.
Normally, that's what markets are for. It is about how good a company is at creating value for the customer. Value, you create customer value in terms of customer experience, again, operationally, but also, and dare I say, most importantly and primarily, through innovation and through the technology and through the total cost of ownership that you enable for your customer. If your product, if your value proposition is superior, then you can attach some more price, and that more price is, in the end, a better economic equation for your customer. You have certainly a valid thought process around that the customer understands.
With that, another thing that we have learned during the last couple of years, is that our pricing processes were not as strong as we needed them to be. We've been very vocal, I think, in the last 6, 9 months, and I feel extremely stronger now in terms of our pricing processes, and we are not done. You know, the ability to price for value-
Mm-hmm.
is a science that we are implementing more so than we had.
Great. Okay. let's see, we have about 5 minutes left. I wanted to ask quickly on ESG, then maybe we'll get to the balance sheet and capital allocation, then we'll open it up for Q&A if we have any time left. ESG is a big deal at Cowen, and it's a big deal because our largest institutional investors demand that we make it a big deal. My sense is that you agree with that, given, you know, I think it was just yesterday, you filed your second-.
Yes
annual ESG report. I'll be honest, I did not have a chance to get through that document, but.
No chat to you, man.
No, not yet. I will. Are there any sort of highlights that maybe you wanna call out? Is this, you know, how much of a box-checking exercise is this for you versus something that you really, you know? Do you guys really care about this?
No, we absolutely do care about that. Yeah, ESG is important to us.
I'm sorry, let me just, I really wanna focus more on the social and the governance first, and then, you know, the environmental piece. Obviously, given your industry, of course, you're gonna care about that, but.
Absolutely. We care about all the three letters of ESG, and you know, it would probably be a lengthy conversation. Really invite everyone to read our ESG 2022 report, that, as you said, was out yesterday. On the social, you know, it stretches multiple aspects. Diversity, certainly in terms of type of education and initiatives, but you look at the composition of my team, actually, is pretty diverse in terms of various dimensions of diversity, and that's something that we foster. We foster a lot of community interaction locally in various countries where we are present. I think we are accelerating in that direction.
It's very, very important for us. I personally believe that diversity is not only important to be as compliant, if you will, in ESG, but a diverse company performs better, performs way better, and that's proven. On the governance side, also very important. Maybe... We can go through many things, but, you know, the fact that the role of the CEO and the chairman are separate...
Yeah
Is an element of governance, strength of governance.
Okay, let's move to the balance sheet, capital allocation. You know, I think if we think about financial leverage on a trailing basis, you still kinda screen a little bit high at about 4.3 x, I think was the number that you called out in March, and that's obviously because that calculation is still including some of the numbers from the supply chain, you know, pressure and so forth. Our model, we project that the company is down at around 3 x net leverage by the end of this year and at 1.5 x levered by the end of 2025. Those are just TD Cowen's numbers. What is your target leverage level? You know, is that lower today than it might have been, let's say, a year or two ago?
Perhaps because we're in a much higher interest rate environment, perhaps because we've lived through the supply chain shock. You have a more conservative view anyway.
Yeah, so traditionally, we've communicated a target between 2x and 3 x for the reasons you mentioned. I'll make a shameless plug for our investor conference coming up in November. We'll be speaking to not only long-term targets for capital allocation, but we may squeeze that range to 2x and 2.5 x. You know, for us, the importance of that leverage is to allow us financial flexibility to do strategic actions, including M&A. We don't see that range as kind of a magic range.
Yeah.
I mean, if we need to float a little bit above it to do a strategic deal, that's fine. If we float a little bit below it, you know, we don't see that as cash burning a hole in our pocket. At the end of the day, that flexibility comes from our ability to generate free cash flow, which has been a huge focus for us, of course, this year, and it will be going forward.
Well, you answered my next couple of questions on the capital allocation there as well. We have just another minute or so left, but I've got a bunch more questions, but are there any questions from the room? Would someone like to jump in with a question before we finish up here? If there are no takers, I'll finish up, but No? Going once, going twice. Okay. Well, since you mentioned M&A, I guess, could you give us a sense for what the potential, you know, targets might be? You know, obviously, E&I was a great fit. We heard really good things about that acquisition from your competitors, who were very worried because it made their jobs harder. You had, you know, that much more product, breadth of product to sell. Are there other sort of...
I'll be honest with you, I wasn't really familiar with Switchgear and Busway before you did that deal. Are there other sort of product areas that maybe we'll read about that you're, you know, reaching into going forward, or how should we think about the M&A opportunity?
Shall I take it real quick? We do not think about something as big as ENI, not in the future, but you never know. You don't know what you don't know, and sometimes opportunities might have their own, their own time. In general, in this moment, we think in terms of technology gap fillers.
Yeah
if anything. As, as we have a very dynamic thermal management environment, a very dynamic power management environment, we didn't talk about that, but there is a lot going on there in terms of transition to renewables, et cetera, that will change, let's say, the landscape on the power side of our business, and hence, the E&I acquisition, very strong. There will be technologies that we may decide to bring into our portfolio. You know, it's very premature. That's the normal activity. It's a portfolio management exercise, and that's what we do quite on a regular basis.
Thank you, gentlemen. We're out of time. I appreciate you being here with us this morning.
Well, thank you so lot.
Best of luck. Thanks.
Thanks a lot.