Podcast: Brian Mushonga & James Hull
I spoke with Brian Mushonga and James Hull, co-Founders of the Caledon Energy Transition Fund, about their backgrounds, partnership and investment strategy.
We used Vertiv Holdings as a case study, touching on its history and development into one of the world’s leading power and thermal management suppliers to the data centre; Chairman David Cote and his track record at Honeywell; its competitive advantages; and its role as a key supplier to the ongoing AI capex boom.
Finally, we discussed how Brian and James approach valuing cyclical stocks; how they overcome anchoring bias; and what could kill the AI capex boom.
The following transcript has been lightly edited with ChatGPT:
Graham Rhodes: Hi everyone, I'm Graham Rhodes, and welcome to the Long River Podcast. I'm really pleased to be joined today by my friends, Brian Meshonga and James Hall from the Caledon Global Energy Transition Fund. This is fun because I know Brian and James personally, and their partnership is quite interesting too. One is based in South Africa, and the other is based in Beijing.
We're going to be talking about their partnership, the fund, and what global energy transition means to them. We'll also do a case study of Vertiv Holdings to illustrate how they think about companies, find ideas, and construct their portfolio. So, Brian and James, welcome to the podcast.
James: Thanks, Graham. It's great to be here.
Brian: Yeah, thanks, Graham. Good to do this.
Before we kick off, we've got to do the usual disclaimer. Nothing we talk about today is investment advice. Everyone should always do their own homework. That's right, isn't it, guys?
James: Yes. I would add that in the future, if you're listening to this, remember that events and things can change. Definitely do your own work and don't rely on this for investment advice.
Okay, so obvious question: How did a guy in Hong Kong meet a guy in South Africa, and how did a guy in South Africa form a partnership with a guy in Beijing? Let's tell that story first. Brian, I think you and I met a couple of years ago through the China Tech Buzz community. You're originally from Zimbabwe. How did you end up in South Africa, and how did you start looking at Chinese stocks?
Brian: Yeah, I grew up in Zimbabwe, came to South Africa to attend university in Cape Town, started working here, and have been here since leaving university, except for a three-year stint in London. I started my career as an actuary, then moved to the markets, joined an investment bank, and became a research analyst covering insurance companies. I left that in 2016 to join an investment holding company investing predominantly in unlisted companies. But once you've been in the markets, you've got the markets in your blood. At some point, I got interested in looking at Chinese companies that were attractively valued. There was an enigma about them; the growth prospects seemed quite strong, and it was a mystery I wanted to unravel. So that was the initial attraction to Chinese stocks.
And James, you used to be a co-host of the China Tech Investor Podcast. I remember just how much I enjoyed your on-the-ground commentary, feedback, and insights. I reached out to you once, and you replied within minutes while walking your dog in Beijing. You called me and were so open and sharing your story. Can you tell us a little about yourself? How did you begin your journey from America to Beijing, and how did you become interested in investing?
James: I got interested in investing through my father when I was very young. I started trading my own account pretty early, as a teenager. I went to university to study economics and finance. After my junior year, I got a job as an intern at a mortgage-backed security CDO hedge fund. They wanted to hire me when I graduated, so I thought, great, senior year and I already have a job lined up. That was 2007. However, at some point, they said they were firing half the staff and couldn’t hire me. So, I decided to travel abroad and went to China. I had met Jim Rogers when I was 14, and he gave me the sage advice to learn Chinese and go to China. I listened to half of it, so when I got to China, I didn’t know Chinese. I learned it myself. By 2010, I joined a hydroelectric utility listed on the New York Stock Exchange, worked there for about four years, and then joined a central SOE in finance doing equity investment and project financing in solar and wind projects in China. So, I have quite a bit of background in energy. In 2017, I left, and I was managing my own money and started doing it for some friends and family. Brian and I met, I think, also through the podcast, and we just started chatting on Twitter and spoke on the phone several times. That’s how I got to China.
From memory, it must have been a long dog walk because we talked for about an hour and a half before you made it home.
James: Yes, I remember that. I no longer live in that community, but I remember that talk well. I was throwing the ball for my dog while we chatted. One of the fun things about doing a podcast is you get inbound connections. I'm sure you get that now with your podcast. I was really thankful you reached out, and so we started a connection and a friendship.
One of those inbound connections was Brian. You guys talked about stocks, found a common connection, and formed Caledon. Tell us about the fund and especially about the strategy. I'd love to know what energy transition means to you and how you identified that as a theme.
Brian: We have a broad definition of the energy transition. We think globally there are changes in the way that energy is consumed and generated. We look at the whole gamut of changes happening in the energy space. In terms of strategy, a lot of our work involves looking at value chains across the energy space. The key is identifying bottlenecks in those value chains because that's where we believe good, sustainable returns can be generated from an investment point of view.
James: To add to that, energy transition means a lot of things, including the carbon transition and the climate change aspect. For us, it means this collective global desire to upgrade our grids, create more energy sources, and ensure energy security. We also look at the use cases of energy, as Brian mentioned. It involves both the supply, such as the grid and construction of new power projects, and the demand and utilization of that energy.
You touched on it a bit, but can you elaborate on what kinds of industries and companies this theme covers?
Brian: It covers a wide range. In terms of sub-themes, electrification is huge, whether it's electrification of mobility, like EVs, or electrification of industry, heating, and cooling. We also look at the enablers of the energy transition. By enablers, I mean construction companies involved in building EV factories, battery factories, and data centers. So, it's quite a wide spread of companies that we examine and select.
James: To add to that, in terms of sectors, our fund right now, in mid-May, is about 56 percent industrials, which includes electrical equipment, engineering, and construction.
And I get the sense that you have some enormous tailwinds behind you, given deglobalization and the industrial policies being launched around the world, such as the Inflation Reduction Act in America.
James: Yes, there's the Inflation Reduction Act. The UK is looking at something similar, though I can't recall the name. There's also the CHIPS Act in the U.S., and many countries are considering or implementing similar policies. So, there's a major tailwind. Deglobalization is a hard word to define, but there's definitely a realization around supply chain security that emerged from COVID-19 and the various geopolitical events creating more volatility in trade. These events, whether in the Middle East or Russia-Ukraine, necessitate a global focus. Governments need to ensure stable electricity supply because we often take it for granted. When Brian and I started talking years ago, he told me about the situation in South Africa with forced brownouts and what people were doing to cope. If you're a government official, you can't let the lights go off, at least not for very long.
Brian: Yeah, just to add to that in terms of tailwinds, the rethinking around supply chains, leading to things like reshoring, means that the demand for electricity is going up in places like North America at the same time as we’re trying to electrify a whole bunch of things. Meanwhile, AI workloads are taking off, which places additional demand on the grids. So, we see strong tailwinds around this theme globally.
And I want to dig into that a little bit, Brian, because one of the things that really caught my attention and impressed me the most about you is how good you are at ferreting out companies. I'm sitting here in Hong Kong, watching Chinese markets for most of my professional career, and all the way over in South Africa, you're finding little companies in the supply chain I've never heard of with amazing financials and great prospects. I really admire that. How do you do it? Is it something really simple? Maybe share some of your secret sauce with me.
Brian: Yeah, I can't say there's a secret sauce, but I have a level of curiosity about how things work and trying to connect the dots. I would attribute it mostly to curiosity. It’s difficult to explain, but networks, curiosity, and connecting dots usually lead you to all sorts of interesting and fascinating companies.
James: Yeah, and just keep digging, right? Find one interesting company, then look at who their suppliers are, who their customers are. Working with you for a while, I've seen you naturally go down those paths.
My experience is colored by what I've seen in China. When I think about the energy transition, it's been a profoundly deflationary experience. When you think about how radically the price of solar panels has fallen over the last 15 years, or the price of electric vehicles, the whole goal is to get to scale and drive prices down. I wonder, as a general question, if that is usually good for equity returns or not.
James: Usually, when we talk about the reduction in cost, we think about the levelized cost of energy (LCOE) and how it relates to fossil fuels. Renewables are now competitive with fossil fuels on this basis. Renewables are all upfront capex; you’re not buying and replenishing fuel. Renewables do require maintenance, but it's much less, and you’re not at the whim of commodity markets. Because renewables are a manufactured good, there are efficiency improvements and scale economies. There’s also environmental relaxation since a lot of the supply chain moved to China, where environmental policies are not as strict as in Europe and the U.S. This contributes to cost reductions as well.
For equity returns, you have to look at the company, its position, its value to customers, and whether it has pricing power. With the bottleneck concept, when a company is the only supplier or has unique traits, they might have pricing power. Whether they use it or not is different, but they might have it, which can be good for equity returns, especially with the large revenue tailwinds from all the policies and capex going into energy, grid buildouts, grid upgrades, and data centers.
While I have you both here, what's it like working as a partnership between South Africa and mainland China? How do you guys do it? How do you divide the work, stay connected, and make investment decisions?
James: We're always chatting. We have persistent chat groups separated into different categories, so we don't have everything in one big chat group. We share ideas, discuss what we find, ask questions, and help each other push that creativity and curiosity. It's great that we have electricity and can have these conversations now. Brian, would you add anything?
Brian: Particularly post-COVID, the concept of remote working and connecting as you and I have done has become a lot more commonplace. The fact that we have these chat groups and always keep the conversations going is quite beneficial. We have recorded texts of what we're thinking and discussing, so it's easier to go back and review. If you have a lightbulb moment, you can revisit stuff you've talked about before. The time difference can be an issue, but it's not a big deal.
Good thing coffee's so cheap in Beijing then.
James: Right? I can stay up later. Having different time zones allows us to go off on our own, do some searching and digging, and then come back to share our findings and insights. It can be quite helpful. If we're always working at the same time, those moments of curiosity might be more difficult to find.
I'm curious, having recently been through this myself, when did you guys commit? When did you know it was the right time to leave your day jobs and launch the fund?
James: We talked about setting up this fund for a while, but we decided to launch in January this year. There's a whole process that goes on before that. Maybe Brian can talk about that. We think the time is now to do this. This theme, this thesis we have, we think it's going to play out for quite a long time.
Brian: Yeah, the timing to launch the fund was right. We could have done it earlier, but there's a process in getting things going. From what we see around the world, whether it's the Inflation Reduction Act or the Europeans looking at the North Sea as a massive energy source, the timing couldn't have been better. We think this is something that's going to be topical for many years to come. The time was right to launch the fund.
You mentioned that you like bottleneck companies and that a large portion of your assets right now are in industrials, which makes the case study you suggested perfect. So, we're going to talk about Vertiv Holdings. My first question is, what does the company do, and how did you hear about it?
Brian: Vertiv is a provider of power and thermal management solutions for data centers, with significant market share in both segments. I've always been interested in data centers, as you know, Graham. When we started talking about energy transition, one aspect of data centers is that they consume a lot of energy. Naturally, I was attracted to look at the architecture and what goes on inside a data center to see how it fits with the theme. When you look at it, two critical functions inside the data center, apart from the computer itself, are power supply and thermal cooling. Finding a company with a leading market share in both segments was a lightbulb moment. That's how we came across Vertiv.
James: To add, power management includes batteries and backup power to ensure that if the electricity coming into the data center or towards the rack is a little off, or there's some kind of frequency problem, it's fixed before it goes to the rack. Thermal management is about cooling, making sure the chips don't overheat, and ensuring the heat is dispersed and removed from the chip, usually ejected out of the data center.
I'm sure most listeners know the answer, but pretend I'm a three-year-old. Why is it important to keep computers cool?
Brian: You've got transistors in semiconductors in a data center. As they switch on and off billions of times a second, they generate heat. If you don't remove that heat, it will negatively affect chip performance and could lead to potential failure. Thermal cooling is crucial to maintaining optimal performance in a data center.
James: If you've ever built your own PC and forget to put the thermal paste on your CPU before starting it up, the thermal paste goes between the CPU and the heat sink. Without it, your CPU could fry, and its lifespan will significantly decline.
Maybe this is a good time to ask about the history of Vertiv, as I think that will illustrate just how this is the right company at the right moment. Could one of you walk us through the history of Vertiv?
Brian: Vertiv's journey began in 1946 with a company called Capital Refrigeration Industries, founded by Ralph Liebert. Eventually, the company took on his name and became known as the Liebert Corporation. Liebert pioneered room air conditioning systems to remove heat from computer architecture. In the 1980s, Liebert was acquired by Emerson Electric, where it remained for the next 30 years until it was sold to a private equity firm. It then rebranded to Vertiv. In 2020, Vertiv went public through a SPAC and became the listed company we know today.
So basically, you have this company specializing in building air conditioning units and cooling systems for computers since the 1960s. It was held by a large conglomerate, sold to a private equity firm, and then taken public through a SPAC in 2020.
Brian: Exactly.
Maybe this is a good point to tell us about Chairman David Cote, his background, and why you think he might be interested in the company.
Brian: David Cote was previously the CEO of Honeywell and did very well there. Before that, he was an executive at GE. At Honeywell, one of the things David Cote became famous for, apart from significant stock price appreciation, was his operational approach. In a nutshell, he focused on keeping costs low and flat while scaling the company in terms of revenues. When you have that combination, your margins expand, and you see the benefits on the bottom line.
If you look at Vertiv's position now, it faces a similar situation. Due to these demand tailwinds, there's an opportunity for massive revenue expansion. However, you have to execute at scale while keeping a tight lid on costs. David Cote has firsthand experience with this at Honeywell, and we think Vertiv can replicate a similar performance. The Vertiv operating system is patterned on similar successes at Honeywell. So, this is certainly not David Cote's first rodeo.
Okay. Before we talk about the major tailwinds, I want to dive a bit more into the details of the company. If I understand correctly, it has 75% revenue exposure to data centers. What's the split between new sales versus aftermarket versus maintenance?
Brian: The split is roughly 3 to 1, new sales versus aftermarket and service.
Who are Vertiv's major customers and partners? Can you tell us a little about the geographic breakdown of their sales?
Brian: The major customers of Vertiv are the usual suspects in the data center world: the hyperscalers like Amazon, Equinix, Google, and Digital Realty. In terms of partnerships, Vertiv collaborates with leading chip and server providers, including Nvidia, Intel, Dell, HPE, and Cisco. Geographically, the Americas constitute about 56% of revenue, with APAC and EMEA each accounting for roughly 22%.
I'm curious. I'm not an engineer, but I'd love to hear your thoughts. As we said before, this company has been building cooling systems for computers since the 1960s. Why hasn't this product been commoditized? In other words, where's the moat?
Brian: The moat lies in the specialized knowledge and expertise required to design and maintain these systems. Cooling and power management in data centers are critical to their performance and reliability. Vertiv has a long history and deep expertise in this niche. They have established relationships with key players in the industry and have a reputation for reliability and innovation. This combination of experience, established relationships, and a track record of reliability creates a significant barrier to entry for competitors.
James: So, where's the moat? I think there are aspects of the product that might be commoditized, for example, if you're buying a PC or something simple for personal or business use. But their customers are building highly important critical infrastructure for their tech businesses. For hyperscalers, a global presence is absolutely important since they have data centers in the U.S., on both coasts, and in many other countries in Asia and Europe. They need to work with partners that have people on the ground who can help them. Vertiv provides a one-stop-shop product portfolio, which is crucial because data centers typically want consistent architecture inside. They don't want one row to have one provider's products and another row to have something else, as that would be overly complex from an operational standpoint.
Additionally, Vertiv has excellent technology and innovation, and they invest heavily in this area, which strengthens their entrenched relationships. For example, Nvidia mentioned Vertiv as a key partner in their last earnings call. Vertiv also has global support with trained service professionals all around the world. If you have a problem in Singapore, they can have someone there very quickly, sometimes even co-located.
So, the moats are the entrenched relationships and the cutting-edge tech in data centers, which are constantly evolving. Nvidia is always developing new chips, and the cooling and power requirements for those chips change accordingly. The architecture of how those chips are put together needs to work with the cooling systems. It's like building Legos; you want your pieces to fit together the right way. That’s essentially a summary of their moat.
Can you tell me about the industry structure? Who are the competitors? Any idea of relative market shares?
James: The competitive landscape in thermal management is divided into two categories: inside the rack and outside the rack. Vertiv is the only vendor in the industry with a complete product portfolio covering air cooling and every technology within liquid cooling, including immersion and direct-to-chip. Vertiv has about 32% market share in this area. The next closest competitor is Stulz, with 16%, and Schneider Electric with about 5%. The rest, about 47%, is divided among various other companies. These numbers are from 2019 but should be similar now.
In power management, the market is even more fragmented. Vertiv has 16%, Schneider Electric around 8-9%, and Eaton about 7%. Again, these are 2019 numbers, but Vertiv remains quite high in market share.
And what's the sales cycle like, and how are sales made? Are they bidding for entire data centers, or are they selling by the rack, by the server? Tell me about that.
Brian: Vertiv engages with customers through multiple channels, including direct sales and a network of channel partners. The sales cycle, particularly for complex projects, is quite extensive, typically lasting months and more likely years. This involves a thorough assessment of a client's needs, designing a tailored solution, presenting the proposal, and negotiating the contract. Building a data center involves securing a site, obtaining permits, and ensuring power supply or connection to the grid, which takes quite a while. So, the sales process for Vertiv is indeed quite long.
James: I think they can start as soon as their customer has a site and has secured power. Basically, when they have the land and power source secured, design and construction can start moving, and Vertiv will be involved from there.
Brian: Given Vertiv's comprehensive portfolio of products across power and thermal management solutions, they often get to bid for a significant portion of what's available in those segments.
James: Right, and it's nice for the customer to deal with one provider instead of having to coordinate multiple groups. It makes the process smoother.
Does Vertiv make its own components and products, or have they largely outsourced their manufacturing to third parties?
Brian: There's an element of both. The company talks about the scale of its procurements, indicating that for certain commoditized elements, they rely on third parties. However, they also do a lot themselves, especially given the R&D that goes into thermal management, for example.
Let's get to the exciting part of the investment case. It's May 2024, and we've just gone through the first quarter earnings for big tech. Each of these companies, maybe barring Apple, has announced a massive increase in their CapEx plans, not just for this year but for the foreseeable future. This is driven by a competitive rush to be ready for generative AI. Tell us about that and how Vertiv fits in. What will this step change mean for demand for their products?
James: All these big tech companies, these hyperscalers, have announced plans to invest heavily and increase CapEx into their data centers. A big part of this is driven by AI advancements, like ChatGPT. Every six months or so, we see significant updates, whether it's improved translations, video, or audiovisual capabilities. These companies are in a race to develop the killer app, and to do that, they must invest in their infrastructure.
They need to build the compute power and infrastructure to support these advancements. This trend is likely to continue for a long time, especially as GPUs keep improving. Consequently, more data centers will be needed, and these companies have the cash to spend on these investments.
Vertiv is essentially a picks-and-shovels play in this scenario. GPUs generate significantly more heat and use more power than CPUs. Large language models require thousands, sometimes tens of thousands, of chips connected together, which generates a substantial amount of heat and requires precise power management. Vertiv provides the critical cooling and power solutions necessary to manage this heat and power, ensuring everything runs smoothly.
Cooling and power management isn't a set-and-forget kind of thing. It requires ongoing monitoring and adjustments. Most chips have heat sensors, allowing for real-time adjustments. As cooling systems improve, there might even be opportunities to retrofit older systems. If you can get chips to run cooler, you can potentially increase their power and extend their useful life.
It sounds like modern data centers are much more densely designed than traditional ones, which presumably generates more heat. This must create a greater need to manage that heat, which is another tailwind for Vertiv, right?
Brian: Yes, definitely. In traditional data centers, you could manage with air cooling. However, as data centers evolve to handle AI workloads and use more GPUs, the amount of heat produced increases. In these denser architectures, air cooling alone is no longer sufficient, and liquid cooling becomes essential.
We're moving towards a world where direct-to-chip liquid cooling is more common. As this happens, Vertiv's market for thermal management expands significantly. Even with direct-to-chip cooling, you still need comprehensive cooling solutions. So, as liquid cooling becomes more prevalent, Vertiv's share of the market for thermal management grows.
How quickly can Vertiv keep pace with demand?
Brian: That's the million-dollar question. It ties back to our earlier discussion about the executive chairman, David Cote, and his experience as CEO of Honeywell. One thing we saw there was his ability to scale operations. We think that experience will be beneficial at Vertiv. Scaling and executing at scale is definitely a priority for the company.
James: Yeah, they mention having 25 percent of their capacity freed up for growth and for handling demand fluctuations. It’s not a perfectly smooth line, but they are building scalability into their expansion plans. This is a good sign. They probably see the market needs even more clearly than we do. They can't afford to miss out because they’re overstretched.
But specifically, what does Vertiv need to do to grow? Do they need to invest in more factories? Hire more people? Do their suppliers need to invest in more factories?
Brian: I think it’s a combination of all of the above. In addition, they have the balance sheet flexibility to expand inorganically through acquisitions of companies with complementary product portfolios or new technology. They’re in a good position to do this, especially given their entrenched relationships with their clients.
We don't have a lot of data for Vertiv because it only went public in 2020. So perhaps this question isn't entirely fair, but why have its historic returns on capital been so low, despite high fixed asset and inventory turnover?
Brian: In the recent past, particularly during COVID, the company was caught off guard by supply chain issues and cost inflation. They didn’t do a good enough job of passing those inflationary increases on to their customers. However, since the start of 2023, with the change in management, the performance has improved. If you look at the quarterly results since then, the company has consistently under-promised and over-delivered, which is a reversal of the previous trend.
So, if you look at Vertiv’s past performance, you might see a different company from what it is now. With the current demand tailwinds and improved operational execution, we believe the company will deliver strong results going forward.
Yeah, and I think that's perfectly captured by the increase in the operating margin from 4 percent in 2022 to the 22 percent that the sell side is forecasting for 2026. So is that simply an improvement from scale, or are they adjusting pricing?
James: Part of it, yeah. Part of it is 2022 marked the low, right? That was the low in their gross margin, which was only about 28%. By 2023, it had already recovered to 35%, so there's, let's call it 7 percentage points there. Estimates project it growing to 38.7%, maybe 39%, so you're getting another 3 or 4 percentage points there. Additionally, SG&A was quite high, much higher as a percentage of revenue, about 20.7%. Consensus estimates predict about a 3 and a half percentage point improvement, bringing it to about 14.5% to 15% in total. Then there's a little bit more, so I see consensus for 2026 at 20% operating margins, which is what management guides. They say they'll reach 20% plus. So they're not saying it's a ceiling, but they say they don't know what it will be. They'll need to reassess based on circumstances. But, just to tally it up, the remaining 1.5 percentage points come from other capex items, like amortization of intangibles, which I think in 2022 was a little higher due to their ENI acquisition in 2021.
So, the numbers make sense. Part of the gross margin increase has already happened, and they foresee some more happening as they scale and gain scale economies in their products. With high demand for their products, they can run the machines a little more. They're currently at 2 shifts a day, so they could go to 3 if necessary, but they're at 2. I think fixed costs will stay where they are, and then revenues ramping up significantly will get them there.
Again, this might not be fair because it's a limited history, but why has free cash flow consistently been so much lower than earnings?
Brian: Given the short financial history of the company coincided with the COVID period, where, because of supply chain issues, there was a significant increase in working capital as companies built up inventory, and Vertiv was no exception to that. But if you look at management's plans in terms of free cash flow conversion, they're giving guidance of around 95 to 100 percent free cash flow conversion, which I think is pretty strong and which I think they can deliver on.
What do they plan to do with that cash?
Brian: They've earmarked about $3 billion over the next three years for buybacks, and we already saw in the first quarter of 2024 the company doing significant buybacks. If I recall correctly, the buybacks they did in the first quarter of this year were north of $600 million. So buybacks are a big part of what they're planning. They've given guidance that, should the right opportunity from an M&A perspective come up, they would look at that. But I suppose that's more opportunistic than anything else.
This has been a great introduction to the business, and I want to take the conversation in a different direction to think about Vertiv as an investment. And really, I want to know how you guys think about investing in a cyclical business. My first question would be, how long do you think the investment boom can go on for?
Brian: This investment boom is going to be sustained for quite a while. We're talking, I think, the next sort of at least three to five years. We think we are at the beginning of a new paradigm in computing with this generative AI. If I just look at some of the tools that I've historically used and how much better they've become because of generative AI in a short period of time, one can imagine what this looks like three, four, five years down the line, right? And an implication of that is I think big tech companies, a lot of companies will have to invest in generative AI just so that they don't fall behind. When you look at it from that point of view, I think these tailwinds are more structural in nature. And we think that actually, for a lot of companies and Vertiv may be one of those companies that have traditionally been cyclical, they face structural demand drivers for the foreseeable future. When that sort of boom ends is anyone's guess, but I think it's here for a while.
Thinking about it differently, can anyone actually find one of these killer apps to justify the investment? Or can the US power supply keep pace with all this construction? Do you have a view on the latter question?
James: So, the power supply is actually another part of our thesis. Because that needs to be built out as well. And there are other companies that do that, that we own today, that will benefit and are already benefiting from that.
So I think there are two sides to it, right? There's the power supply keeping pace. There are the utilities who, especially in the US, have seen energy requirements of industry go down a little bit as a lot of things were offshored or outsourced in the last couple of decades.
At the same time, we're using way more electricity in our refrigerators than we were 20 years ago, not to mention cell phones. Constantly, we always have some sort of electric device with us that we're using. And so they, but because the demand that they were seeing was not really increasing, they didn't really have to do much with their supply investments.
And now that they see there's the reshoring, friend-shoring, reindustrialization kind of thing going on, there's obviously a CapEx boom in data centers. Everyone's like, "Whoa, this is a lot more than we were planning." And the electricity grid is not something you can just go and start building things. It has to be planned, balanced, and connected in a way that's not going to disrupt the other parts of the grid. So it's like a giant machine basically. And it needs to be well thought out, planned.
So they have processes and they do it, but none of this is super fast. It's not move fast, break things Facebook-style. So the power supply will, I think, keep pace. But because also the chips are starting to focus a little more on being power efficient, but what might often happen if you read Vaclav Smil and you study the history of energy use is even though things get more efficient, we end up using more energy.
So they might just use more chips. They might say, "Oh, we got a thousand Watts or something from the grid. We got a gigawatt data center. Now the chips are more efficient. We can put even more chips in there." It might not actually lower the overall energy consumption, which is what matters for the grid, right?
What makes this different from the build-out of fiber and telecommunications networks in the 1990s?
Brian: So I think part of the difference, Graham, is that you were talking earlier about killer apps. We are already using generative AI, right? The ability to actually leverage off the infrastructure that's being built is already there. I use Gemini pretty much on a daily basis. If I look at something such as Third Bridge, the productivity I get out of Third Bridge has gone up in leaps and bounds as a result of generative AI. So I think we are already seeing the utility and the benefits of this new technology, whereas I think that at the time of the build-out of fiber in the late nineties, early 2000s, the ability to use that infrastructure maybe wasn't as well developed as it should have been. You didn't have an iPhone back in 2000.
James: They don't want to have someone else destroy their bread and butter, right? If they have to, they don't want to be the disrupted. They want to be the disruptor. Yeah. So it's a life-and-death, literally a bit. They have to do it. I think also back then there was quite a bit of, and I could be completely wrong about this, but I think there was some fraud and accounting fraud going on. And some of these companies were not as cashflow generative as they may be. We're trying to make it out to be. And it ended up being a boon to the internet. But there was a lot of capital destruction. But it ended up basically meaning that the cost of those assets, because they had to, they went bankrupt or they were bought out of bankruptcy. There's a lot of capital destruction and ended up lowering the cost overall and being net beneficial. And when you spread, look beyond those companies.
So I think that's, it's different now because these companies that are investing in this, they have very good businesses with strong cash flows. So I feel like that's a big difference. I think the CEO mentioned that recently as well. And I agree with him. If they weren't so cash generative and they were borrowing a lot to do this, the sustainability of it would definitely come into question.
Yeah, I do take the lesson that you pointed out, that the users of these assets in the nineties benefited greatly, but the builders of them went bust. So that kind of leads me to my last question here, which is how to think about this as an investment. Again, it's a cyclical company, but we could be, or we already have entered what could be a super cycle. So I'm just curious to know, like, how do you think about valuation? What's the best way to value a company like this at this point in time?
Brian: Look, I think from a valuation perspective, the market always puts a premium on certainty, right? So certainty, consistency, and visibility. And we think that Vertiv is well on its way to deliver on all three fronts, as far as its revenues and earnings are concerned. Assuming continued secular tailwinds, which we assume consistent execution by management, normalizing supply chains, which we've already seen in Q1 2024, and strong incremental margins that James was talking about earlier, we see a credible pathway to $6.50 in EPS over the next few years, bearing in mind that the company did $1 EPS or $1 Uni seven in FY 2023.
And companies that demonstrate consistently high single-digit revenue growth, double-digit EPS growth, and expanding market share in industrial companies often trade within a 25 to 30 PE range, or if you want to take it from an EV/EBITDA multiple, 18 to 23 multiple. We think that should Vertiv deliver successfully on the targets that management have set for 2027, investor confidence will grow, and that will potentially justify a premium valuation similar to what we've seen for some industrial companies.
Okay. And just on a human level, the stock is up I think more than 10X from its low. How do you get over the anchoring bias there?
James: If you're trying to think about it, you just have to size it appropriately. If you're overly concentrated, it's very hard to buy something like this just because of that anchoring bias. It's hard to avoid. But I think, the 2022 low, there was a lot going on. Their earnings were down, their cashflow was down, had a lot of working capital issues related to accounts receivable, the COVID disruptions, and all that, the bullwhip effect and everything. So I think you need to look forward and what Brian's just talked about there. They can trade at that multiple or near that multiple in the future. And if management can execute and they're somewhat conservative. There was a small comment in a recent, I think it was the B of A conference, they kind of alluded to the possibility that they have an internal estimate and an external estimate. The external one is a little bit more conservative. We could be undershooting here a little bit if things go right. But, I don't know if people usually mention risks. There's a lot of risks, there's risks that we keep an eye out for when we're managing our own and other people's money. And we want to make sure that they execute and things are going well, and we're keeping an eye on those risks.
Brian: One of the things I'm finding with these companies with sustained sort of massive tailwinds is the operating leverage that's going on. Yeah. It's difficult to model, right? So if Vertiv can actually get it right on pricing power, keeping a tight lid on expenses, the impact on the bottom line can be massive, right? And it's very difficult, I think, for analysts to model massive gains in operating leverage, right? If there's anything I've learned quite recently at some of these companies, as long as they're moving in the right direction. So keep checking the order book. So hopefully, touch wood, Vertiv can do the same thing. Execute well, and it all drops to the bottom line.
James: One of the risks is if everyone at the company can look at their results, right? They're a public company. So they can start asking for higher salaries, or there'll probably be an increase in dilution. But if they're generating cash and they're able to offset that with share buybacks, that would be ideal. Actually, more than offset would be ideal. But at least offset a little bit because their employees. You have a currency as a public company and not using it would be silly. Just paying everyone a higher salary doesn't help your margins, and you can offset that with some SBC.
Actually, that's a good question. So if you think back to COVID, there was a shortage of software engineers, and we saw all the big tech companies, their margins just go to hell in a handbasket. Is there a similar thing going on in with the industrials in America, people to build the data centers, engineers to design this stuff, are there enough of them or are you seeing this stuff as well?
James: There's definitely shortages, and I think what ends up happening is, in order to secure the contracts, I think they have to make sure they have the labor. And so it ends up being who has the labor wins. And so there's a number of companies in our portfolio right now that have that kind of situation. High demand for a scarce resource—labor and expertise—will usually mean higher prices. And so how the companies manage that is important. And you can see how they've managed it in the past. We talk about all this stuff when we look at these industrials, and we've been, the funny thing is, we start, we launched in January. We've been working on this for years. We've been researching these companies for a while.
Yeah, maybe you can elaborate specifically on the frequency or the cadence of technological change. You hinted at it earlier in the switch from air cooling to liquid cooling. But I've read in a couple of places that this transition might open a window for competition.
James: At the end of the day, Vertiv still has a pretty high market share in power management. So they're gonna, they have a foot in the door there, no matter what. And then they have the technology and they bought, I think it's called CoolTerra. So they have the offering and if they're talking to their partners, Nvidia and their other partners, we presume they're going to be a step ahead of the competition.
Brian: Yeah, I'd agree with that. I think if you look at the scale of the big tech companies, the hyperscalers, their global footprint, their capex plans, it's very difficult for those companies to be dealing with small startups, right? They want to deal with a global partner. So I think that puts Vertiv in a very good position to acquire. I think startups that have pretty competitive technology, right? It's just gonna be very difficult for a startup to break through and start dealing with an Amazon or a Microsoft. It's much easier when they bring their technology through iv. So I think those intrinsic relationships that IV has with its partners puts it in a very good position and aligns to grow and then change that position even further.
Okay, Brian, James. It’s been absolutely fascinating talking with you guys. I'm really grateful for what you shared today. I've certainly learned a lot. If anyone listening to this wants to get in touch with you guys, what's the best way to be in contact?
James: Go ahead and email us. I'm james.hull@caledonip.com. And on Twitter, I'm @jameshullx.
Brian: And I'm brian@caledonip.com. On Twitter, my handle is @mushonga_brian
Hey, guys, thank you very much for joining the podcast. It's a pleasure to have you. And once again, I've learned a lot and I'm very grateful. Thanks.
James: Thanks, Graham. Thanks for having us.
Brian: Thanks, Graham. Yeah, it was a pleasure. Enjoyed this.