Transcript:

Caroline WoodsWe’ve heard about the brains of AI, the chips and the software, but the conversation is shifting toward the body, the massive infrastructure and power required to keep those brains alive. Joining us to break down the physical side of AI is Matt Sallee, Head of Investments at Tortoise Capital. Matt, thanks so much for being here at the desk.

Matt SalleeThanks for having me, Caroline.

Caroline WoodsSo as we think about portfolio construction in 2026, do we need both brains and body? We’re talking about the body, obviously, but do we need both in our portfolios?

Matt SalleeI think we need both. Most people have the brains already. If you own S&P 500, you’re going to own plenty of Nvidia or other chip makers. That’s why we differentiate with our fund and we’re more focused on the body as you describe it, which is funny. That’s how we talk about it as well. This is the physical infrastructure that’s required for AI to to be provided. So if you think about the servers, the racks, the big thing obviously is electricity. That’s the gating factor. Everybody used to think that it was. It did used to be chips. Now it’s all about electricity. And then things like cooling and the other physical infrastructure. So that’s really where we’re focused. And when you hear about hyperscalers, hyperscalers spending $400 billion, our portfolio are the companies that they’re spending that money with.

Caroline WoodsOkay. So talk to us about what some of those companies actually are.

Matt SalleeSo Modine Manufacturing as an example there. This is funny. It’s 100 plus year old company. They used to make car radiators. They’ve pivoted. They obviously pivoted their business a lot over the years. Now they’re focused on liquid cooling. That’s also one of the big constraints is, you know, these chips obviously run really hot. They use a lot of electricity. So that requires a lot of cooling. You can do direct air cooling. There’s a number of different types of cooling, but, liquid is becoming the preferred method. And that’s what Modine does as an example.

Caroline WoodsAnd you can get exposure to Modine through your tortoise AI Infrastructure energy fund. What are some of your other biggest holdings?

Matt SalleeYeah. So in TCI that’s what we’re focused on is the infrastructure. Biggest holdings are companies like third of another liquid cooling name. Arista Networks is a networking as a name would suggest. Really think about it as, kind of the traffic cops of, of, getting data around, Dell Technologies is another one where, you know, obviously have pivoted. You know, I always historically thought it was a PC maker. Now they’re making AI server racks and again, this is a company that still trades at a wildly discounted valuation, even though it had a good run in 2025. And so as they pivot their business that we expect them to continue to rearrange so those are some examples. We also have, big exposure to some of the independent power producers. So I talked about electricity. There’s electricity is the gating factor. And these companies are procuring energy as quickly as possible. One example is Vistra Energy. They just made a $4 billion acquisition yesterday of gas plants five gigawatts, which to put that in perspective, that’s enough to power New York City on an average day. So they get it for a really good price. It’s located in the, PJM market, which is the northeast, as well as the Ercot market, which is Texas. Those are the two arguably most attractive markets in the country. And, so it’s really all about speed to market. And that’s what they’re doing, is acquiring any assets they can because it takes a while to build a new power plant.

Caroline WoodsIt’s interesting because I saw in your notes that you called electricity the new oil and an AI driven economy, but what does that actually mean for investors?

Matt SalleeSo I can’t take credit for that. I love it, but my fellow portfolio manager coined that phrase. But, so what it means is electricity is really the future of energy. Oil will continue to be an important part of the global energy mix, but it will lose market share in our belief and what will certainly gain market share and has been its electricity. So if you look at electricity in our country, the last two decades has been really been stagnant. In terms of demand. We’re at the precipice of that really changing and growing at a significant clip for for electricity. I mean, we’re talking three and a half, 4%, which, you know, and expected to double by 2050. So the need for electricity is huge. Spending on electricity last year, capital spending just blew my mind. I’m a I’m an energy nerd, though.

Caroline WoodsI’m an energy nerd.

Matt SalleeI had, spending globally on electricity was greater than capital spending on oil, which is kind of wild to think about. And that’s where the bulk of the CapEx dollars we think are going to be spent going forward.

Caroline WoodsAs as we think about the brains versus the body. You mentioned that investors probably already have exposure to the brains because of the S&P 500. But given the fact that the S&P is so tech heavy, would you say that investors are underestimating the energy and infrastructure opportunity?

Matt Sallee100%. That’s one of our, you know, biggest, call a complaint. But, energy is 3% of the S&P. It’s, you know, most people have an allocation between 0 and 5%. If you look at the earnings weight of energy in the S&P, it’s over 10%. If you look at the weighting of energy in the economy, like just how important it is to our economy, it’s over 15% of GDP. So energy is really underrepresented in most portfolios. And underappreciated. And it’s one part of the market that actually trades at a pretty reasonable valuation discount of valuation versus the other market is is at all time highs.

Caroline WoodsSo you can’t get exposure through a fund tracking just the S&P 500. Why get it through a thematic ETF like yours versus just stock picking some of those names that you mentioned.

Matt SalleeSo we have a portfolio of about 50 names. And it really takes a balance of the what. We break it down into energy infrastructure digital infrastructure and the data centers themselves. So it’s really a packaging of all parts that have to happen to provide AI really outside of the hyperscalers. So these are the companies benefiting from the hyperscalers. And what we’re doing, what makes us unique is is actively managed. And we’re looking for, you know, kind of diamonds in the rough, if you will. So as an example, when, one set of companies that we’ve invested in and have done quite well are the bitcoin miners and you’re thinking had a Bitcoin miners fit in, well, they procured power to make bitcoin several years ago. And what they’re doing is and effectively what these are data centers that are producing Bitcoin. What they’re doing is they have to beef them up a b o and add redundancy, but then they’re able to provide high performance compute services or hosting services. And they’re then under long term contracts with fixed rates rather than being exposed to bitcoin prices. And you know, have massive volatility in the earnings. What’s that lead to a significant rerating or higher. Higher back cap multiple. So you know finding diamonds in the rough like that as I think some of our special sauce and why our fund is unique.

Caroline WoodsWho is the bigger winner in the AI powered race? Is it nuclear energy? Is it natural gas?

Matt SalleeNatural gas is going to be the big winner. Simply because. And we’re big fans of both. Natural gas is the big winner because we have tons of it. It’s really cheap. And that’s the quickest way to get to market in terms of new power. Now nuclear is also going to play a role. There’s three different reactors that have been mothballed that are being brought back to service, and those will have a massive impact. They’re big. You know, one of them as an example is two gigawatt facility, which again, going back to, like I said, average in New York. Consumption to about five gigawatts. So these are big plants. So we’ll bring those three back. That can happen relatively quickly in terms of building a new nuclear plant. We’ll see. It takes. Some time. Right? It will take a lot of time. The last time it was built, was in Georgia, and it was, you know, ten years over ten years delayed and about, you know, two times over budget. So, you know, I think the regulatory environment has gotten easier, but, you know, it’s still it’s a risky proposition to build new, new large scale nuclear. A lot of companies are focused very hard on small modular reactors. And, that’s something again, it’s probably ten years or at least five years out, if not ten. The technology works. It’s a matter of making it economical. And we’re believers. We just think it’s going to be a little while. We actually created a new ETF, the Tortoise Nuclear Renaissance Fund te nuke truck, that’s invested in nuclear power plants, large scale and small scale, as well as uranium enrichment, you know, kind of the whole value chain, which it’s become a national security imperative to reassure nuclear, you know, the nuclear supply chain, which we really have relied on for and sometimes not, you know, not real friendly countries. So we bring that back, reassure the supply chain. We see some pretty good opportunities there, too, with the nuclear.

Caroline WoodsYou mentioned the regulatory landscape getting a bit easier. So just to wrap things up, I’m curious, are there things like regulation constraints, supply constraints, tech issues that could kind of derail the the AI, energy and infrastructure thesis.

Matt SalleeIf you will? I think the biggest thing the the hyperscalers are super well funded. They produce tons of free cash flow. They have eight, triple A credit rating. So never worried about them having the the wherewithal to spend. And I think they’re going to spend. The question is adoption. I think that’s a pretty modest risk in terms of I don’t know about you, but, pick your service that you like. I use I, you know, on a daily basis. Now, whereas even just a year ago, it was pretty rare that I would use it. And I think corporations or businesses are just now adopting and realizing the efficiencies that can be gained. And potentially, you know, hires avoided or just make your business run better. I think we’re in the very early stages of that. The other thing that could derail it is it talks about, electricity is the gating factor, and it’s not like you can throw up a power plant overnight. So, you know, that’s a challenge. And then, you know, just in terms of permitting, the biggest thing that what we have seen some significant improvement in terms of building pipelines. So there’s an enormous amount of natural gas and it’s next to free in the Marcellus Shale in the northeast. And then, you know, northeast really needs more natural gas. So, specifically to produce electricity. So, we’ve seen a couple pipelines. The Williams companies, had proposed a number of years ago that they kind of gave up on because they couldn’t get the, permits. Those pipelines are now moving forward. So significant improvements, something really concrete, we can point to in terms of permitting a permit that will make a real difference in reliability and allowing I to.

Caroline WoodsOkay, we’ll leave it there. Thanks so much for shedding some light on this space. Really appreciate it.

Matt SalleeThanks, Caroline.

Caroline WoodsIt’s Matt Sallee of Tortoise Capital.