BONUS EPISODE: America's $41 Trillion Debt Crisis + Can AI Save the U.S. Economy? | ATM Live

Never. Never.

And I'm Mike Novogratz.

Yeah, that's our chant in the mornings, to wake ourselves up. So, Michael, let's go right into the macro situation. U.S. rates: some people think rates are going a lot lower. They think the war is ending very quickly. Other people think rates are going a lot higher. Where are rates going?

Well, to start, we're in a bear market in interest rates. So, every investor and every trader should write that on your wall: bear market in interest rates. What happens when you don't have the direction right? You're like, 'Oh, it's time to buy now. Oh, it's time to buy now.' It doesn't mean rates are going to go one direction non-stop. But this is a bear market. We're through all the moving averages. It trades like a bear market. We have an enormous amount of supply from the government, right? $41 trillion of debt, even though Scott Bessent is going to try to tweak the curve. That didn't go well. Right, his old mentor said, 'What are you doing?' But forgetting that, SpaceX alone next year wants to raise $100 billion. And so there's like a trillion dollars plus of data center capex debt that's going to get raised. And so every rally in the 10-year and the 30-year needs to be sold.

And so really the trade is the curve, right? The long end, at best, if oil comes down and we have some weakness and unemployment comes up, at best the long end, the 10-year, can maybe get to 4.70 and then it'll bounce, bounce, bounce. And so the new range of the 10-year: 4.70 to 6%, 4.80 to 5.80, you pick it, right? Or higher. The front end, more tricky. You've got a new Fed governor, Kevin Warsh, and no matter what he said to Trump to get the seat, Kevin Warsh is a very persuasive man.

He said, 'Right now I'd be cutting interest rates.' And I believed he— I believe he believed it when he said it, and facts changed, and he's changed his mind, and he's going to hike interest rates again. Maybe not the next meeting, but the following meeting. He cares a lot about his reputation, a lot. He's waited for this job a long time. He's really disciplined, and so I honestly think he's exactly the guy we want at the helm right now. And so, listen, I think he's going to raise rates a little bit, but I do think the AI revolution is real. There's a couple years left, or at least a year left, of capex inflation.

I run a company called Galaxy. We're building one of the biggest data centers in America, and I can say without a shadow of a doubt that there's a lot of inflation in the data center building business. But I can also look at the new Muse app, or the new Insight, if Anthropic's got that one, it's even better, and say, whoa, the deflation is coming, right? The job replacement is coming. And so if we start seeing job losses, because we really haven't yet, right? But we've seen them in young coders, the only place AI has really hurt. But if we start seeing job losses at law firms and job losses at accounting firms and secretaries being fired because the personal assistant is so damn good, you know, you're going to see that front end come back down. And so the trade that every portfolio manager should have is the yield curve steepener, twos-thirties, which are at plus 80 basis points. Now, the difference between the two-year and the 30-year will end up sometime in the next three to four years at 400. And when we were young—

Yeah. Yeah.

No, but that— but I only got halfway through macro.

When Trump got elected, and you know, I was on the other side of the fence on that election, I was at least thrilled that it looked like Warsh would be his guy and Scott Bessent was coming in, and Scott was telling a story that was needed, that we were going to try to bend the curve of debt to GDP, and we were going to do it by having growth— 3% growth, 3% fiscal deficit, right? We were at six at that point. We were going to get it down to three, and $3 energy was his three, three and three, and we're at 7% budget deficit now, right? Scott so far has been really unsuccessful. Not all his fault. Congress does what Congress wants to do. The president does what the president wants to do. And instead of cutting spending, writing, growing ourselves out of the debt, we've increased spending by trillions. And so we're now at a 7% fiscal deficit.

$2 trillion is one reason the economy is doing so well. Spend an extra $2 trillion, the economy feels all right. And I don't see that shifting yet. The bond market is telling— it should be telling all of Congress. It's certainly telling Scott Bessent, 'Danger, Robinson, danger.' Right? It's feeling like— and it's not just the U.S. bond market, the European bond market, the Japanese bond market. And so this great bubble that we've had for most of our professional lives, right? This government bond market bubble is being deflated, and we're not sure where that stops.

Office of Management and Budget.

Yeah. David Stockman was the young firebrand head of the OMB for Reagan. And at that point we had a thing that the whole world was talking about, or the whole country was— the twin deficits. And we had a 5% fiscal deficit and a 5% trade deficit. And that was red alert, red alert, red alert. And so back then, we used to try to tax— 20% was the federal tax take, and you wouldn't want to spend 20%. And if that got out of whack, we were only taxing 15% and still spending 20%, there was your deficit. What was interesting is just the amount of nervousness around that, of focus on changing that, like that was a really big deal. And what has happened? We balanced the budget one year since I've been alive, really, it was 1999 with, you know, Clinton and Gingrich, and you guys also have to thank the other Bill— not Bill Clinton, but Bill Gates, right? You had this internet bubble and this productivity gain and all those capital gains, and that's how we balanced the budget. But that, you know, starting with 9/11 and then in '08, people kind of lost grip of that, and today we're still taxing at 20%, but we're at a 27% spend rate, and we won't— we'll never get the 27% back to 20 because Social Security, Medicare, these entitlements are just a bigger portion of it. But we've got to stop it from 27, at least get it to 24.

Anthony's a good Catholic, so we should all— we should all— well, he at least is a Catholic. I don't—

We should all say a prayer that AI works. And what I mean by that is, I was talking to— not Dario, but the other co-founder of Anthropic, he's their economist, and he was saying, 'Listen, we've run a ton of models on job loss,' and he said, 'There are some models we're running that literally have the U.S. economy growing at 6% to 7% with inflation down at 2.' Like, and I was like, 'Okay, I'm praying for that.' We need AI to work for this country to get ourselves out of this.

And good jobs, right? And wage growth, right?

Listen, like we need a couple of things. We need immigration to come back to the country— not the immigration we had during the Biden administration, which was unchecked and destabilizing— but this country needs at least 2 million immigrants a year to have the economics work. And right now, we're at net zero. So that needs to change. But we need this AI productivity miracle to happen. We need growth better than 2% to 3% to grow ourselves out of this debt problem. And if we get it, if we get 5% growth and 2% inflation, the $41 trillion of debt won't be so bad.

Yes. So, we're really a holding company with two businesses now. So, I'll start with the data center business. We have a big data center business. We're, you know, currently building the first 800 megawatts, which people don't understand how big that is. I went down there, and it's much bigger in real life than in the picture. But we're about an eighth finished of the first campus, and it was daunting. And so just on our first 200 megawatts, to put it in perspective, right, there were what looked like railroad cars on top of this big shack. And I went and opened the railroad car up, and there was a gigantic diesel generator, looked like a giant motor the size of the railroad car, that we had bought from Caterpillar. And that's 1/8 of the first data center of the first campus. And so when you start looking at the capex build, you understand why the economy is chugging along, because to build our whole, you know, 1.6 gigawatts that we now have approval for— and those generators only get used if the grid goes down, so you're hoping they never even get turned on, which is kind of crazy. And so the first thing I noticed was the capex build in America, because we're one of many data centers, is big, real, is going to take a huge amount of capital, but is going to continue to support the economy. We are lucky that we're building a data center in a state that is very pro-business and, you know, with sparse population around us, who we've already seen the income level rise significantly in the county that we're in. You know, we've been good community members. We built a pool, and, you know, we're doing all kinds of things to support the community.

And they recycle the water.

You can think of it like the radiator in your— we use like a huge amount of water that we will flush through the system once every two years. So for two years you use no water, and then once every two years you'll flush it out and you'll restock the pipes. The noise that people talk about— I was down there, you know, they use the same chillers that— because data centers really are giant air conditioners. I was like, 'We just built a huge air conditioner, didn't we?' And they're like, 'That's it, right?' So these chillers, the same chillers that you find on top of high schools for the air conditioning unit, not really that noisy. So a lot of the fears of the physical data centers are just overblown, but they do represent AI, and AI is scaring people. And, you know, if you're an X-ray technician, if you're a— like, every job can go to, 'What's going to happen to me?' And so both parties are working and need to work on, like, their AI policy, and it's happening so fast it hasn't been developed yet. And so what does that mean? That means, what's the tax policy? There's going to be an AI dividend, right? You think about just AI at its core. You're taking historical knowledge, all of human history's knowledge, and all of our data, and you're putting it into a box, and you're dumping energy and computer chips, and it creates knowledge. So that's it. That's how it works. Even like Dario Amodei said, we don't really understand much more than that, you know. And so if it's using all of our knowledge, by definition there should be some dividend to humanity. That's the political framing that any smart politician is going to use and say, 'Okay, what does that look like? Does that look like a token tax? Does that look like, as Trump has suggested a few times, let me take stakes in each of these companies? Does it look like a tax on energy used?' What smart guys need to do together now is figure out what's the most efficient tax and how do we then distribute it to the people that are most nervous.

Bullish on that side.

You know, crypto is at this great inflection point. So, let's split crypto into two halves. There's Bitcoin as this digital gold for this next generation. And it had been kind of meh for most of the year. And the moment Scott Bessent said, 'I think I'm going to mess with the Treasury market in the 30 years,' everybody woke up and said, '$41 trillion in debt, and I don't have a plan.' And so crypto went from 60,000 to 77,000 today. And Bitcoin itself feels better, and it's brought the rest of crypto up, and I don't think it's going back down. It might go down to 70, but we now feel like we've survived yet another Bitcoin bear market. And so I'm bullish Bitcoin. The other side of crypto has always been, we're going to build this decentralized new plumbing— centralized and decentralized new plumbing for the financial services and quite frankly, and for consumer services. And that had been just a story. People used that story to whip up energy and then create these tokens that everybody traded around like gambling tokens. And we're now at the point where the rubber is meeting the road, and that new technology, that infrastructure of blockchains and tokenization, is starting to happen. And so if it's Bank of New York, we've partnered at Galaxy with Bank of New York, but we're partnering with lots of TradFi institutions. I would tell you every single big TradFi institution has a major tokenization, a major blockchain effort. And so the idea that there'll be crypto assets and traditional assets in two years won't be an idea. Every asset will have a tokenized form. Some assets will have an old-school form and a tokenized form. And so that the big bull market and what we're working on at Galaxy— a good chunk of what we do now is digital infrastructure.

We are 100% in a bull market for crypto infrastructure. Every single traditional bank is going to spend more money next year than they spent this year on building infrastructure. We're in a bull market for tokenization. You will see more and more assets tokenized.

And I think we're in a bull market for Bitcoin. I just don't see us getting out of this debt problem anytime soon. I pray we do. I would put my Bitcoin back in half if we could fix this structural deficit in the United States, right? But I think it's going to be much harder.