If they're running away, we're going to go in. I mean, our job is to be contrarians. We're very busy in Europe. Very busy. Find the freight trains in your life and get on them instead of in front of them. And the freight train for a 25-year-old today is AI. You know, it's really how do you think? Are you asking the right questions?
Barry Sternlicht is founder and CEO of Starwood Capital Group. Welcome, Barry. Pleasure to have you here.
Thank you. Nice to be here.
You've just had an interesting panel. Tell us about it.
Well, the panel was about the future infrastructure of what's the best infrastructure asset class in 2030. And I think because everyone's fixated on AI, and of course we ourselves are a very large data center builder, we could say data centers, but data centers are just one end of the infrastructure. You have to have the power, obviously. The power is the fuel, the oil that creates the data center. And I think people talk about power. There's in many countries that we operate, the United States in particular, we have gas here, they have oil and gas, that's not the issue. It's actually getting the power to the data center, so it's transmission. And our panel was Mohamed Alabbar from Emaar, and he was talking about residential also, that you know, that you need housing. This is a global housing shortage, and I think that was a good pivot because I think everyone's just going to talk about data centers, it's the flavor of the hour. And I think residential is so underinvested. And it's not just, you know, Emaar does very high-end properties primarily in 20 countries, but also Dubai. You need affordable housing. The world needs housing. In the United States, we have almost a four, now a five million housing shortage. And you're building housing shortages in Spain and in Sweden and all over the world here. But it's affordable housing, and you have a rising population that needs to get housed. So they're going to have to figure out that puzzle, especially in high-cost labor markets, because it's not just the materials, it's also the labor. So nobody's figured out, for example, modular construction. It's not worked at scale yet. And maybe AI will help figure out the processes to bring cost down. I know the design times will become very fast. I've seen, we invested in companies that do AI home design, and you can do it, what used to take you six months will take you three minutes, and they can cost it out. So I think you're going to see enormous benefits there. But at the end of the day, labor in high-cost markets, that'll be a challenge. So it'll be fascinating. And the other thing I thought about the panel was the discussion about, the moderator who's another from a network said, you know, they're going to spend a trillion dollars on data centers by 2030. Not if the revenue model doesn't show up. I mean, people are, yes, there's such a frenzy now, and we enjoy the frenzy. We're doing really well in our data center business, but it will slow down if the revenues don't show up.
So you're being cautious in how you look at it?
I think you should be more cautious than that. I do think the road is up, but there will be bumps. People will fail and it will shake confidence and it will sort itself out. And in our industry, the data center industry, we're probably building $20 billion of data centers today around the US and Europe and also Australia and through another company we control in Japan and Malaysia and probably India, ESR. It's a big money manager based in Hong Kong. I think you're going to see, we have to get out of them. They're so big. Who buys them? Like we can build them and we can build them at really good yields, very attractive versus other asset classes in real estate, but who's the buyer? And I think we're all asking ourselves the question of like who's going to buy that $38 billion data center, that the new, the Stargates are $30 billion, 18, 20 billion. Even for PIF and other investors, these are big checks to own one of these.
I do remember in Miami, the FII Priority, you said the data centers are the world's most valuable real estate.
They're not the world's most valuable real estate, but they were, you could build them to the highest yields on cost of any real estate asset class. And the spread between what we can build them for and what we think we can sell them for is the highest. So I think that made them and brought a lot of real estate guys who couldn't build apartments. Office doesn't work. You know, we don't need another shopping mall in the United States. So a lot of real estate guys retooled themselves into data center guys. And then they met the infrastructure guys and we clash, you know, so we're both doing it. We have about 200 people worldwide working just in data centers today. But we do everything. We do hotels, you know, we do residential, we do everything. Condos, we do everything globally.
I hear that a lot of your business is based in Europe, but here at the FII, we hear many people very pessimistic about Europe and investing into Europe. What's your experience?
