people장 홈📈 Stock Masters주식 유튜브를 Articles로 읽고 Author의 예측을 실제 가격으로 확인합니다 people장 Board 홈요청하기

한국어 · English

Why Ray Dalio Says the AI Boom Is a Bubble (And Why He's Not Selling)

the0stack · 2026-08-16 · 유튜브 시청 134 · Board Views 0 · International

International

유튜브 썸네일

원문 보기

다듬은 글 보기

Ray Dalio just said the quiet part out loud. The man who saw 2008 coming, who made money while everyone else lost theirs, looked at the AI boom and used one word, bubble. But, here's what the headlines missed. Dalio didn't say sell. He said something weirder, that the bubble and the future can both be true at the same time.

So, today what Dalio actually sees when this thing pops, and the one number he's watching that almost nobody's talking about. Here's a hint. It's not the stock price. First, why should you care what one guy thinks? Because this isn't just any guy. Ray Dalio built Bridgewater into the biggest hedge fund on the planet.

Hundreds of billions. And back in 2007, while the banks were high-fiving each other, his team was quietly telling clients the whole system was about to crack. Then 2008 happened. The banks burned. Bridgewater's main fund went up. So, when Dalio says bubble, you don't scroll past it. You ask him to show his work.

And here's why his warning is worth more than a hot take. Dalio doesn't guess. He built a machine, a model of how debt cycles work. It goes like this. Cheap money shows up. Everyone borrows to buy the hot thing. Prices climb, so people borrow even more against the gains. For a while it feels unstoppable.

Then the debt gets too big to pay back. Lending dries up. And the whole thing runs in reverse. He mapped that same pattern across 50 crises and 100 years of history. 2008 fitted exactly. And when he looks at how this AI boom is being paid for, he sees the same shape forming. Here's the work. Dalio ran the numbers and said today's AI stocks look like the late 1990s.

The dot-com run-up. Prices flying away from what the companies actually earned. Everyone piling in because it only goes up until it doesn't. He pointed at how much of the whole market is now just a handful of AI names. Get those few wrong and you don't have a dip. You have a hole. But, and this is the part everyone skips.

He did not say the technology is fake. In 1999, the internet was real. It did change everything. The companies just got about a decade ahead of themselves. Pets.com died. Amazon didn't. That's Dalio's actual warning. Not AI is fake. It's AI is real and the price still isn't. So, what's he actually staring at?

Two numbers. First, concentration. A tiny handful of AI-linked stocks now drive most of the market's gains. Call it the Magnificent Seven. When seven companies basically are the market, the market's whole future rides on their bets paying off. Second, price. Those stocks aren't priced for things go fine.

They're priced for things go perfectly forever. Dalio's own bubble gauge measures exactly that gap. How much of the price is hard earnings versus pure hope? Right now, that gauge is deep in the red. Okay. The question everyone is literally Googling, "When does it pop?" Bubbles don't pop because somebody rings a bell.

They pop when the money stops circling. And in AI, the money is doing something strange. It's moving in a circle. Watch this. Nvidia invests billions into OpenAI. OpenAI turns around and spends it renting computing power from Oracle and from a company called CoreWeave. Those companies use that money to buy chips from Nvidia.

Round and round and round. Add it all up and it's something like $800 billion in deals. A big chunk of it, the same money, counted again and again on everyone's books. On the way up, it looks incredible. Every company's spending becomes another company's record revenue. The The problem is what that same circle does on the way down.

Now, look closer at one strand of that circle because it's the part that makes analysts uneasy. Nvidia doesn't just sell chips into this loop. It invests in the companies that buy its chips. It committed up to a hundred billion dollars to Open AI. Money Open AI then spends on computing power that runs on Nvidia chips.

Read that twice. The chipmaker is funding its own customers so they can afford to buy from it. That can be smart aggressive strategy. It can also quietly inflate everyone's numbers. Nvidia books the sale, the customer books the investment. Until someone in the chain actually needs cash and finds out it was mostly just circulating.

Here's where it gets fragile. To build all this, the data centers, the power, the chips, the companies in the middle didn't use spare cash. They borrowed heavily. Oracle took on a mountain of debt to build data centers for Open AI. Billions in fresh bonds. And Coreweave, look at this one. 15 billion in debt against roughly 4 billion in revenue.

About four times what the company earns and hundreds of millions every quarter just to cover the interest. Now, the important question, where did all that lending money come from? Mostly not the banks. It came from private credit, pension funds, insurance companies, the boring safe money. Yeah, some of your retirement money is quietly in the trade chasing a slightly better return.

So, Dalio's real signal was never Nvidia's stock price. It's the debt. The first crack isn't a crash on the news. It's quiet. A funding round that doesn't close. A loan that can't roll over. A credit fund that suddenly tells investors they can't pull their cash out. That's the domino. Watch the debt, not the ticker.

Okay. So, what does it actually look like when this pops? Because it won't be a movie style crash. It'll be a chain reaction, and it starts in a place no one's watching. Step one, money stays expensive. Rates don't fall fast enough. Step two, one of those debt-loaded cloud companies goes to refinance.

And the lenders either say no or say yes at a rate that eats it alive. Step three, it can't cover the interest. It sells hardware at fire sale prices or it defaults. Step four, the private credit fund that lent it the money takes the loss. Its investors panic and ask for their cash back. But the fund is illiquid, so it slams the gate shut.

Step five, now the safe money is trapped. Every other lender gets scared, and the lending just stops. And that's the kill shot. Because the companies in that circle needed constant new money to keep the deals going. When it stops, the circular revenue stops with it. And it turns out a lot of it was never really there.

That's how a financing bubble ends. Quietly in the credit market, months before the stock market gets the memo. Now, before you sell everything and go buy a bunker, the other side. Because this is Dalio's whole point. Both things are true. This is not the year 2000. Back then, companies with no product and no revenue were worth billions.

Today, Nvidia is selling every single chip it can make to real customers paying real cash. The biggest spenders, Microsoft, Google, Amazon, are some of the most profitable businesses in history. And a lot of that AI demand is already locked in under multi-year contracts. So, this isn't a house of cards.

It's more like a real building, a solid one with way too much borrowed money stacked on just a few of the floors. The tech doesn't have to fail for the finance to wobble. That's the whole trap. And I have to steal them in the other side, hard, because it's real. The giants bankrolling most of this, they can actually afford it.

Microsoft, Google, Amazon, Meta, throw off tens of billions in cold, hard, free cash flow every quarter. If AI spending disappointed, they'd slash it and survive. Bruised, not buried. That's the real difference from the bubbles Dalio keeps naming. 1929 was crushing leverage with no safety net. 2000 was companies with no revenue at all.

Today, the core of this thing, the big platforms, have real customers, real profit, and real cash. The danger was never the core. It's the edge. The levered middle that borrowed everything to keep up. So, forget trying to time it. Here's what to actually watch. Three signals in plain sight. One, refinancing.

The next time a big AI data center player has to roll over its debt, watch the rate. If it spikes, that's the fever. Two, the private credit funds. Watch for any of them limiting withdrawals. A fund that won't let investors out is the smoke. Three, the deals themselves. Watch for a circular deal that gets quietly walked back or shrunk.

That's the money starting to notice. None of those are stock prices. They're all debt. That's the whole point. So, here's what I actually think. Dalio isn't calling the end of the world. He's telling you to stop staring at the wrong thing. Everyone's watching Nvidia's stock price like it's the smoke alarm.

It isn't. Watch the debt. Watch the levered middle of that circle. The core weaves, the neo clouds, the ones who borrowed to exist. Because the first one to break is never the strongest company. It's the one that owes the most. And if you want an exact date, stop. There isn't one. Watch for the signal instead.

The first big AI loan that quietly can't get refinanced. When that headline shows up, the clock has started, not a day before. That's the whole game. The man who called 2028 says AI is a bubble and the real thing in the same breath. The hype crowd only hears it's real and ignores the debt. The doom crowd only hears bubble and ignores the technology.

Dalio's telling you to hold both. The boom is real. The financing is fragile. Watch the money, not the noise. If that made the loop finally click, subscribe. That's the entire point of this channel, the signal without the hype. One last thing. We talked about the money tonight, but there's a harder limit coming that money can't buy its way around.

Power, as in electricity. That's the next one. This is the zero stack. See you there.

#헤지펀드#엔비디아#현금#금리

← List으로
공유페이스북밴드

이 Articles에 대해 요청(숨김) 수정 요청

Comments 0