Data center REITs are companies that buy and rent out the real estate that houses data centers and processors for companies like Google, Facebook, and Uber.
In this week's episode of Industry Focus: Financials, hosts Michael Douglass and Motley Fool contributor Jordan Wathen walk listeners through what investors need to know about data center REITs before diving into the industry. Find out how REITs in general work, and how data center REITs are unique; some of the most important risks facing the data center REIT industry in the next few years and decades; some of the most compelling reasons investors might want to take a look at this industry for their portfolio; and more.
A full transcript follows the video.
This video was recorded on Sept. 25, 2017.
Michael Douglass: Welcome to Industry Focus, the podcast that dives into a different sector of the stock market every day. It's Monday, September 25, and we're talking data REITs. I'm your host, Michael Douglass, and I'm joined by fool.com's Jordan Wathen. Jordan, welcome back to the show!
Jordan Wathen: Thanks! Good to be here.
Douglass: Good to have you! For listeners, I'm filling in for Gaby Lapera while she's out of the office. Let's hop right in. The first thing with data center REITs is, probably the first question most people ask is, what are data centers? So, Jordan, let's turn to that first, and we'll get to REITs and why we're doing this on Financials.
Wathen: To start, data centers are basically really big, secure warehouses where companies store racks of servers or computers that store tons of data and move tons of data. These computers and servers also do the thinking behind what we see as a really basic thing as an end-user. If you go to Google, for instance, and type in "cheap flights" -- all of that computing, Google has to go through its host of data that it has on the internet and see, what are the best sites to return for this? That computing power happens somewhere far away from you in a data center.
Douglass: Yeah. What's interesting about this is, longtime listeners will know we divide Industry Focus into five different sectors each week -- Financials, Consumer Goods, Tech, Healthcare, and Energy/Industrials. But the fact of the matter is, there's a lot of crossover between these different groups, and data center REITs are one of those areas where you could argue that it's tech, or it's financials. But the thing is, these REITs, or real estate investment trusts, are, at their core, balance-sheet-driven financials companies. So, that's why we're covering this in Financials. These guys basically buy out this real estate, and then lease out the computing power to the Googles and the Facebooks of the world.
Wathen: Right. It's kind of complicated, but if you think about what a REIT really does, they basically buy real estate and hard, long-lived assets --
Wathen: Very expensive stuff. And they turn it into an easy monthly payment. Which, if you think about it, it's basically what banks do. You buy a house with a mortgage, they turn it into an easy monthly payment. Or a company wants to buy a new location for a pharmacy, a Walgreens wants to buy a new location, they don't want to hold the real estate, so they pass it off to a REIT. And the REIT owns the real estate, and Walgreens just pays them a monthly payment. So, for data centers, think, an Uber, they don't want to build the data centers and hold onto them. They would rather just lease the capacity from a data center REIT like Digital Realty Trust.
Douglass: Yeah. Particularly if that REIT already has unused capacity that the tech company could use when it's scaling -- because, when you think about an Uber or a Facebook, or really any of these companies that have rocketed in terms of their volume and their transactions, and as a result, their data capacity that they need, they can't easily build new capacity. What they really need is someone who already has that capacity that they can lease so they can scale up very quickly.
Wathen: Right. A lot of data centers talk about how bringing a new one online is basically a two- or three-year project. And Uber, since we already brought it up, just last year, it basically doubled in size and the number of trips and the amount of revenue it was doing, the amount of bookings it was doing. So, a company like Uber really can't build out as fast as it can build its underlying core business.
Douglass: Yeah, absolutely. Let's turn a little bit to what's in it for the REITs. For so many reasons, this is good for the tech companies. They don't have to own these big, expensive assets. They don't have to do all of the scaling up with all the upfront capital expenditure they have to do. What's in it for the REITs on the flip side?
Wathen: On the flip side, what the REITs get is a long-term contracted cash flow. They basically get a promise from the company from, say, an Uber, we'll just use that as an example, that they will pay $X per month for a certain amount of computing capacity or space or whatever, and they will pay that for X number of years. It's basically like lending money to the tech industry, in a way.
Douglass: And it's like an annuity for the REITs, which is a good deal for them. Now, let's talk a little bit about thinking about REITs as a whole, and these data center REITs in particular. If you're thinking about investing in this space, what sets one apart from another?
Wathen: One of the biggest differentiators is scale. There's multiple ways to measure scale. The first way is financial scale. You have to be pretty big to issue a bond. If you're a REIT who wants to finance real estate with a bond issuance, you need to be able to issue $200 million or $300 million at a time. That's a whole lot of money. You can't just do that overnight. You can't just do that as an upstart. And I guess there's also, to an extent, a geographical component. If you as a customer of a data center REIT can go to one data center REIT and say, "Look, I need to be in 50 countries tomorrow," and that data center REIT can do that for you because it already has the space and all these different data centers around the world, that's an advantage, too, I'd say.
Douglass: Yeah. I think the other big advantage is one across REITs. Real estate investment trusts as a whole are required by law to pay out almost all of their earnings as dividends. So, as a result, when they're trying to raise money and take money and spend it on something, like, for example, buying a new warehouse, or in this case, a new data center, they have to take out debt to do so, or they have to issue equity. And if they're going to take out debt, one of the big things is that they need to be able to get that debt at a relatively low interest rate. If you are, let's say it costs you, I'm making up this number, $10 million a year in interest to have a mortgage on a new data center that you just purchased, you can't be leasing it out at less than that. You're not going to make a dollar if you do that. So, you have to be leasing out at more than that. So, low cost of borrowing is important, and that's why size really matters. Frankly, if you're going to lend to a business that has a market cap of $2 billion versus one that has a market cap of $10 billion, you're going to prefer the $10 billion business because it's just bigger, and that probably means it's more stable. All other things being equal, of course. And the other piece of that is strong balance sheet. If you have a REIT that is relatively underleveraged, it's going to be more able to take out that kind of debt, and therefore get that debt at a lower interest rate so that it can compete more effectively for customers.
Wathen: In some way, it's almost like looking at banks. The bank with the lowest cost deposits wins, generally speaking. They don't have to take on more risk, they don't have to do crazy things, they just have that financing advantage that's so huge. So, you can make a lot of mistakes and still do just fine.
Douglass: [laughs] Right. And we do like businesses that are forgiving in that way. The other piece of it is, as interest rates creep up, and long term, that's going to happen, regardless of what the Fed does in one quarter or another. We're at still near-historically low interest rates, and the Fed has signaled that the plan is to continue to keep stepping that up. That differential between somebody with, let's say, a AAA balance sheet versus somebody with a BBB balance sheet is going to widen. And that's where cost of capital is going to become really important. In a lot of ways, REITs have had it easy for the last few years because capital has been really cheap to get [since] interest rates have been so low. As that gets harder, the operators that have a lower cost of capital are going to see a bigger and bigger advantage.
Wathen: That's a really good point. And that's one of the reasons why one of the biggest differentiators, especially in the financial world, is basically a credit rating. Once you can get an investment-grade credit rating, you can almost have a license, especially right now, right now definitely, a license to borrow money. If you're a junk borrower right now, you still kind of have a license to borrow money, it's not too hard. But when a 2009 comes around, it won't be the junk borrowers of the world that get money on great terms or at all -- it'll be the pristine balance sheet companies, the companies that have the stability to be able to issue long-term financing at a low cost.
Douglass: Yeah, absolutely. Let's step back a little bit and look at the broad scope here for a minute. Looking at the broad trend of tech and data, and very much the trend is toward more data, and that means more storage, and that means data center REITs will benefit. One of the real benefits of investing in data center REITs is, they will benefit because of the broad trend. You don't have to pick the winner in Amazon versus whoever, or in Google versus whoever, or in Facebook versus whoever to benefit, because these REITs will benefit just because everyone is using more data, and therefore needs more storage. So, that's one of the really nice things about investing in trends through REITs. They provide a necessary service. But, let's face it, there's kind of a dark side, a weakness to it, too.
Wathen: Let's think about that. Before the show, I was thinking, what's something that I could draw a corollary to? And I think the most obvious trend you could possibly see ever is that, there will be more need for food in the future as the population goes up. So, theoretically, farmland should be a great asset. But, if you look at the history of farmland, for example, the real money that you made there over the course of time has been, for the most part, from interest rates going down. Even though food is a great growth industry, it's going to go up as more calories are consumed, but it hasn't been a knock-the-lights-out kind of asset because at its core, it's a really simple, pure, boring business. It's a really boring asset.
Douglass: And even more than that, it's an asset in this case that tech companies, who are going to be driving so much of this growth and benefiting so much from this growth, don't want. Which should be a little bit of a sign.
Wathen: You have to think about how these companies add value. At the most basic level, Google and Facebook aren't great investments just because they use data centers, or just because they own data centers. What makes them great is that they're exceptional advertising companies. What drives Facebook's or Google's earnings is how much advertising they can show, and how many people click on them and how much they can charge for it. That's an added value layer on top of the fact that they're using data centers to make this happen. So, I suspect, really, even for these businesses, if the cost of their data centers doubled overnight, it wouldn't really matter because of how much value they add on top of that data moving around.
Douglass: Which does imply that there might be some pricing power for the REITs. Except for the fact that they're competing actively with each other. And many of them share the same clients. So, Facebook can say, "These guys are charging us $2, and these guys are charging us $1.50, so we're going to go more with the $1.50 guys."
Wathen: It's interesting. I wanted to speak to someone who has some knowledge of this, who really understands the most basic level, and to some extent, it can be super commoditized, but there's also geographical balances. So, right now, I was reading a report, Denver apparently has excess capacity in data centers, so it's actually cheaper than other markets. But, if you have someone accessing the data in China, the place in Denver doesn't help you that much. It takes too much time for data to travel through the tubes. So, it might not be the best choice to basically put something out on a grand scale across the world. Real estate is all local. And that goes for data centers, just as it does retail real estate, I think.
Douglass: Absolutely, that makes a lot of sense. Jordan, let's step back to the actual crux of the question. And by the way, listeners, if you ever have a question, and you're curious about having it answered, shoot us an email at email@example.com. Patrick sent us a question about data center REITs, it's why we did this whole episode. So, if you want an episode dedicated to you, just send us an email, firstname.lastname@example.org. I guess, really, Jordan, the key question for us is, what do you think of data center REITs as an investment? Listen, obviously, they're going to be around. These folks are probably not going to be going out of business anytime soon. They've got big sector trends that benefit them. On the flip side, they invest in an asset that may not be as attractive as the actual tech companies. So, as an investor, how do you think about this sector as a whole?
Wathen: As the sector as a whole, it's kind of disinteresting to me. Data centers, there's going to be more data in the future, as you said. I have a hard time understanding who's going to win and who's going to lose. So, when Equinix says, "We have the most pristine asset in Miami, Florida. It's the only way or the best way to connect the United States to Latin America," I hear that, but I don't really know what that means, and I don't understand what that will mean 10 years from now. So, whether or not I want to invest in a stock at a 3% yield based on something I don't really know that well, it's hard for me to wrap my head around. When you're paying a high multiple of earnings, you have to have a lot of certainty. Otherwise, that investment doesn't make sense. I can't get comfortable with a margin of safety with the price where it is. I can't get comfortable with how much I can be wrong about it and still make money.
Douglass: Yeah, I think that's a fair concern. For me, one of my big concerns is that someone will figure out how to disrupt this area. When I'm thinking about industries that are adjacent to tech, a lot of times, they get disrupted as tech folks, some Silicon Valley entrepreneur, or somebody thinks of a different and better way to do things. And I worry that someone will figure out how to store and call up this data in a cheaper fashion, and that will then require these REITs not to go out of business, but to retool. And the cost of that will be so substantial that it'll hamper growth. Plus, frankly, I'm concerned about interest rates growing. REITs as a whole tend to benefit enormously in a low-interest rate environment. They tend to suffer in higher interest-rate environments. And we're looking at a higher interest rate environment long term. To be honest, there are sectors I tend to invest in in real estate. I tend to like healthcare for real estate, in part because it's multiples are lower and the dividend yields are a lot higher. So, I do tend to think there's a little bit more margin for safety there. And also because [with] healthcare, the actual delivery of care isn't being disrupted nearly as much as things like data storage really have that potential to be. So, I tend to think of that as a safer place, where there won't be a potentially enormous, life-changing thing that requires everyone to retool and shut off growth for a couple of years.
Wathen: That's actually a really good point. If we think about capacity as far as data center REITs are concerned, it's based on the amount of electricity you can pump into the building, the amount of computing you can do with a certain amount of electricity. And over time, that goes down, you can do more computing with less electricity. Whereas, if you think within the confines of a traditional REIT, say an apartment REIT, for example, people aren't going to want to live in 100-square-foot units. There's a certain minimum.
Douglass: Maybe it's not 100, but it's something.
Wathen: Yeah. It's something. I've seen studios recently being built that are tiny. But, there's a limit to the minimum. I don't really understand where that is with tech. And I think you're right, I think there's a lot of potential disruption risk there, because you're building long-lived assets. Was this data boom a big thing 20 years ago? No. But these buildings are going to be standing 20 years from now. So, it's hard for me to wrap my head around.
Douglass: So, Patrick, that's our general take on data center REITs. It's interesting, Jordan and I both spent some time looking at the various REITs. I really struggled to find key things that distinguish them outside of the general things with REITs. So, in terms of balance sheet, in terms of credit rating, in terms of dividend yield. Outside of that, it seemed that their cost of electricity production and the sort of things that they were touting they could do were broadly similar. Sure, there were some geographic differences. But it was hard for me to really identify a company that I felt was reasonably moated.
Wathen: Yeah, I think that's fair. One of the things that a lot of companies are talking about now, and let's get in briefly, the things that companies are really talking about now in this space is the ability to interconnect. So, basically, this is really bad, especially for REITs, but think of a data center as a shopping mall. You have your anchor tenants like a Macy's or JCPenney or something, and those bring the people in. In a mall, in a small geographic area, one mall will survive because it'll have the best anchor tenants, and everyone will want to go shop at that mall. So, a data center REIT's pitch now is basically, "Look, we bring together all the awesome clients, and we can connect them directly to one another through one cable right here on site, so that'll bring everyone else to us, too, and we have some pricing power because of that." And that may be true right now, but I don't know how that changes in the future, or how companies end up putting their data in hundreds of different data centers, and whether that edge remains. But, it's kind of an interesting argument, that the interconnection part of the business is the good part of the business. I think that's something to look at if you're following this industry and have more interest in it.
Douglass: Yeah, absolutely. Folks, I think that's it from us for today. If you'd like to get in touch, email us at email@example.com. Questions, comments, ideas -- we love them. Or, tweet us on Twitter @MFIndustryFocus. As always, people on the program may own companies discussed on the show, and The Motley Fool may have formal recommendations for or against stocks mentioned, so don't buy or sell anything based solely on what you hear. For Jordan, I'm Michael Douglass. Thanks for listening and Fool on!
Suzanne Frey, an executive at Alphabet, is a member of The Motley Fool's board of directors. Jordan Wathen has no position in any of the stocks mentioned. Michael Douglass owns shares of Alphabet (C shares), Amazon, and Facebook. The Motley Fool owns shares of and recommends Alphabet (A and C shares), Amazon, Facebook, and Twitter. The Motley Fool recommends Equinix. The Motley Fool has a disclosure policy.