저자: RayDalio · 외국인
요약 보기
버블의 마지막 단계에 진입했다는 주장. 신용시장이 주식시장보다 먼저 경고 신호를 보내며, AI 인프라 관련 대규모 부채와 소수 기업 집중이 취약점. 유동성(금리·글로벌 유동성) 방향이 사이클 지속 여부를 좌우한다고 봄. 다만 타이밍 예측은 어렵고 무조건 현금화는 불리하다는 중립적 결론.
다듬은 글
The final phase of the bubble has begun. I've lived through enough of these cycles. The 1970s, the dot-com crash, 2008, to recognize the pattern at the moment when it begins to repeat itself. And it's happening right now in real time in front of all of us. What I want to do in this video is not to scare you.
I want to do the opposite. I want to give you the simple mental model I use to see these things coming, so you're not the last to know what's happening, like most people are. Because here is the eternal truth. Bubbles don't burst when everyone starts to worry. They explode when the last people who could join have already done so and there is no one left to continue driving the price upwards.
We are closer to that point than most people think. So stick with me, because by the end of this you'll understand exactly where we are in the cycle and, more importantly , what to do about it. Before we continue, I want to be honest with you about something. What I'm about to explain took me decades of studying these cycles , losing money, and rebuilding my thinking from scratch to fully understand it.
Most people pay a lot of money for this type of reasoning. In books, seminars, consulting fees; I'm giving it to you for free. In simple terms, right now . All I ask in return is that you subscribe to this channel . Before you continue, it costs you nothing and means you won't miss the next time I analyze what's really going on behind the headlines.
Alright, let's get down to business. Now, many people will hear the word bubble and immediately assume that I'm telling them to panic, sell everything, and run for the exit. I don't do it. In fact, that's the worst thing you could do, and I'll explain exactly why later in this video. What I'm telling you is much more useful.
Understanding where we are in a cycle is one of the most valuable pieces of information you can have, because every cycle in economic history, without exception, has followed a recognizable pattern. Not an identical pattern, but a recognizable one. Just as winter always follows autumn, although no two winters are exactly alike.
My entire approach to markets has been built around the study of hundreds of these cycles in different countries and centuries. And what I've found is remarkably consistent. The mechanics don't change, only the characters and the technology change. The 1920s had radio and automobiles. The 90s had the internet.
Today we have artificial intelligence. History is always different. The underlying mechanism that drives the boom and the inevitable bust is almost always the same. Let's start with the most basic question because most people get it wrong . What is a bubble? A bubble is not simply a market that has risen a lot.
Prices can rise significantly because something has genuinely become more valuable, and that's a healthy thing. A bubble is something very specific. This is when the price of something rises mainly because people expect it to continue rising, and that expectation is financed with borrowed money. That's all.
That's the whole engine. So you have an idea that sounds true. Homeownership will continue to expand. The internet will change everything. Artificial intelligence will rule the world. And that idea is usually correct. But then debt comes into play and changes everything, because debt allows people to buy today what they can only afford with tomorrow's income.
When enough people and companies start doing that at the same time based on the same story, it creates a self-reinforcing cycle . The price increase makes the story seem true. The fact that the story seems true attracts more loans. More loans push prices even higher. And so it keeps turning over and over again.
And for a while it seems like free money, as if the old rules no longer apply. That feeling, that sense that this time is different, is the most constant warning sign in every bubble in history. I have never seen a bubble form without someone confidently telling me why the normal rules of debt and cash flow no longer apply to this particular case.
And right now you're hearing some version of that phrase about artificial intelligence. That alone should make you pay attention . Not because AI isn't real. It is absolutely real and it will change the economy. But being right about a technology and being right about the investment cycle around that technology are two completely different things.
The internet was real too. Everything changed exactly as people said it would . And the companies that capitalized on that story still caused investors to lose half their money when the debt behind it collapsed. Being first and being right is not the same as being invested at the right price with the right amount of leverage.
I want to give you a simple way to test if something is a bubble, because I use the same checklist myself . Ask yourself, are prices high relative to traditional measures of value? Do people expect today's rapid price increases to continue? Is the purchase being heavily financed with debt? Are buyers acquiring much more than they normally would in some cases, simply for fear of missing out ?
And has speculative buying attracted a wave of new, less experienced participants who weren't there before? When you can answer yes to most of these questions at the same time, it's very likely that you're in a bubble. Regardless of how exciting or real the underlying story is , I invite you to apply this checklist to artificial intelligence today, honestly and without emotion, just as you would analyze any other market.
You might be surprised to see how many boxes you check. Now, this is where I want you to understand the stages, because bubbles don't happen randomly. They almost always follow a sequence. And once you see the sequence, you can't stop watching it. The first stage is real. Something genuinely valuable emerges.
A new technology, a new asset, a new way of doing business, and smart capital comes in first using its own funds because the returns are clearly attractive. This stage is healthy and necessary. This is how capitalism finds good ideas. The second stage is where things begin to change. As the first returns become noticeable, more capital wants to participate.
But the easy and obvious opportunities have already been taken. So, instead of demanding better returns for the risk they take, participants start accepting increasingly worse terms just to be part of the story and continue growing at the same rate. Those leading the boom begin to resort to debt, because debt allows them to grow faster than real income would allow .
This is the stage we are currently immersed in with artificial intelligence. The companies that build the infrastructure, data centers, chips, and power plants needed to operate all of this are taking on debt on a scale that would have been unthinkable just a few years ago, precisely because their cash flow is not keeping pace with how quickly they want to build.
The third stage, which I really want you to focus on , is when the story becomes more important than the numbers. Valuations are no longer based on what a company earns today, but are now based purely on what people believe it could earn in the distant future. And when that happens, the entire structure becomes incredibly sensitive to trust.
As long as everyone continues to believe in the story, the money keeps flowing and the structure remains. But a change in confidence, without the need for a change in the underlying technology, is enough to violently reverse everything. This is the phase I believe we are in now with AI. Not because the technology is fake, but because the funding behind it has become almost entirely dependent on continued belief rather than a steady cash flow.
I have seen this third stage develop so many times that I can almost predict the language that emerges along with it. You start to hear that traditional valuation methods no longer apply to this new category of businesses. You start to hear that the companies that lose the most money are actually the most interesting investments because they invest aggressively in the future.
You start to see huge sums of capital committed years in advance based on projections rather than results. None of this automatically means that a disaster is imminent. Sometimes history really does come true, and the first believers are rewarded immensely . But it does mean that the system has entered a phase where confidence, not fundamentals, does the heavy lifting.
And systems that operate on trust are by definition fragile, because trust can change direction much faster than fundamentals. Here's something that took me years to fully appreciate. And once you understand that, you'll never see the markets the same way again. In reality, there are two markets happening at the same time and they almost never agree in the later stages of a cycle.
There's the stock market, where people buy shares of companies expecting unlimited growth, and the credit market, where people lend money simply expecting to be repaid with interest. These two groups think completely differently because their incentives are completely different. The best possible outcome for a stock investor is enormous.
The action can be multiplied many times. That's why stock market investors are naturally optimistic and willing to ignore risk as long as growth continues. The best possible outcome for a lender, on the other hand, is very limited at most. You get your money back plus a modest interest rate, but your worst-case scenario, if the borrower defaults, is losing everything you lent, since the upswings and downs are very unbalanced for a lender.
Lenders are naturally much more cautious and pay much closer attention to whether the debt can actually be repaid. That's why , historically, the credit market almost always sees problems before the stock market. Lenders are starting to withdraw. Insurance against non-payment is starting to get more expensive.
Borrowing costs are starting to rise quietly. All of this happens discreetly beneath the surface. As stock prices continue to reach new highs and headlines celebrate record growth, the stock market is finally catching up with what the credit market already knew. And by the time ordinary people feel it in their jobs or savings, the cycle is already ending , not beginning.
This exact sequence occurred before the dot-com crash and before the 2008 financial crisis. In both cases, warning signs in the credit market appeared months before stock prices moved and long before the average person felt pain in the real economy. I want you to reflect on that sequence because it tells you something important.
By the time a bubble is obvious to everyone, it's too late to act on that information. The people who protect themselves are those who pay attention to the silent signs, not the noisy headlines. Think about it for a moment from the lender's perspective, because this is the part that most people never stop to consider.
If you are a bank or a bondholder that has lent billions to a company, your only daily job is to ask a question. Will I get this money back? He doesn't care how exciting the growing-up story is. He doesn't care about the size of the total addressable market in 10 years. You care about cash flow, guarantees, and the likelihood of payment.
That's a completely different perspective from the one most retail investors use when they buy stocks excited about the future. And because lenders observe so closely and their risk is so serious if they are wrong, their collective behavior often changes before public sentiment. Observing how easily companies can still borrow and at what cost says much more about the health of a boom than the stock price.
So, let's return to the present, because I think this is truly one of the most important situations I've studied in a long time. The amount of debt being taken on to build artificial intelligence infrastructure right now is enormous. Larger than the debt behind the expansion of telecoms in the dot-com era and larger even than the subprime mortgage market that triggered the 2008 crisis.
That alone doesn't make it a disaster. Debt itself is not the enemy. Debt only becomes dangerous when the income to cover it does not arrive on time or when the cost of borrowing rises faster than expected. And here's the fragile part that I think most people are overlooking. A huge portion of this AI-related debt and spending depends on just a small group of companies that actually generate the revenue to justify it.
If those few companies continue to grow as they have been, this cycle may continue to extend. And if they don't, or if the cost of capital increases before they get there, the whole structure will be tested very quickly. It's a bit like balancing something huge and heavy on a very small and very specific point.
As long as that point remains firm, everything may seem stable, even impressive from the outside. But the margin of error is minimal because there are very few alternative points that can support the same weight . This is exactly the type of concentration risk that appears just before every debt-fueled boom reaches its limits.
Not because the central idea was wrong, but because it was built too much on too narrow a foundation, financed with too much borrowed trust. What makes this moment unusual, even by historical standards, is how interconnected everything has become. Companies that take out loans to build physical infrastructure need a small group of AI developers to keep investing to justify that debt.
These AI developers, in turn, need continuous external capital investment to keep paying for that infrastructure, and that capital needs the confidence that both will continue to grow together. It's a closed cycle. And closed loops can be incredibly efficient when everything works. But they can also transmit tension very quickly if just one link in the chain slows down.
In previous cycles, borrowers and end users of the product used to be millions of independent individuals: home buyers, internet users, and small businesses. That diversification meant that no single failure could bring down the entire structure. Today, an unusually large part of the foundation of this boom depends on the consistent success of a very small number of companies.
That's not a criticism of those companies. It is simply a structural fact worth understanding, because concentrated risk behaves very differently from distributed risk. When things go wrong, this brings me to the point that most people ignore, and which, in my experience, is the most important variable in any cycle like this .
Liquidity. Liquidity simply means how easy and cheap it is to move money through the system. When liquidity is plentiful, interest rates are low, and credit is easy to obtain, investors are patient. They are willing to finance companies that are not yet profitable because they believe they will be rewarded and because there is cheap capital to continue financing the story.
But liquidity is not permanent. It goes up and down depending on the decisions of central banks and the confidence of lenders. And when liquidity contracts even slightly, everything built on borrowed optimism is suddenly exposed. This is exactly what happened in 2000 and again in 2008. Rising interest rates made borrowing more expensive just when the most leveraged parts of the economy could least afford it.
And the story that had justified all that debt collapsed almost overnight. Right now, the direction of interest rates and global liquidity is the most important thing to watch if you want to know how much longer this AI-driven cycle can last. If liquidity remains favorable, this could last longer than skeptics expect, sometimes for years.
If liquidity contracts significantly, the fragile and narrow foundations under this boom will be quickly tested. And that is usually when the final phase of a bubble becomes its end. Actually, there is a simple and almost mechanical way to understand why liquidity matters much more than people realize.
Imagine all the companies that are currently losing money but continue to raise funds because investors believe in their future. That funding is essentially a bet that someone will continue to supply fresh capital until the company is profitable on its own. As long as fresh capital continues to appear at low cost, that bet continues to pay off and the company survives long enough to try to prove its value.
But the moment fresh capital becomes expensive or scarce, that same company can go from prosperous to insolvent surprisingly quickly, even though nothing in its underlying business has changed. That's why many companies that seemed unstoppable during periods of easy money suddenly appear fragile as soon as interest rates rise.
Actually, it was never about their product. It was about the cost and availability of the capital that kept them afloat. So here's my honest take on where we stand. We are no longer at the beginning of this story, and we are probably not at the end either. We are in what I would call the final phase.
The stage where the technology is real and the growth is real, but the funding behind it has become strained, concentrated, and dependent on continued faith rather than cash flow. This is the stage where the credit market usually begins. Whispering warnings before the stock market realizes it, where the final outcome depends largely on two things.
If the small number of companies at the center of this boom can continue to grow to their enormous valuations and if the cost of money remains low enough to give them time to achieve this. Both things could actually happen either way from now on. So that uncertainty is exactly why this moment matters so much.
Um, I also want to be honest about something. I learned it the hard way throughout my career. People who are usually right about a bubble are often wrong, sometimes by years, about exactly when it ends . Markets can be irrational and remain funded much longer than logic would suggest, because as long as new capital keeps coming in, the story can keep going.
I've seen skeptics be right in the end while ruining themselves in the process, simply by trying to calculate the exact moment of the collapse instead of managing their exposure to it. That is an important distinction and it is the difference between intelligently analyzing a cycle and trying to bet on its end.
So what does all this mean for you, who are watching this video trying to decide what to do with your own money? It doesn't mean you should sell everything and hide in cash. History shows that this approach often costs people more than the collapse itself, because nobody rings a bell at the top or at the bottom.
What it does mean is that this is exactly the kind of environment where blindly sticking with what has worked in recent years , without questioning your exposure to leverage and concentration, becomes truly risky. The goal is not to predict the exact day this ends. The goal is to understand the machinery well enough so as not to be taken by surprise when the credit market starts sending out its warnings in the same way it always has.
silently before others notice. Pay attention to how easily the largest AI-related companies can continue to borrow money, and at what cost . Pay attention to whether the handful of companies at the center of this boom continue to grow in line with their enormous valuations. And pay attention to where interest rates and liquidity are headed, because that single variable has decided the outcome of almost every cycle like this throughout history.
Understand the machinery, respect the cycle, and position yourself so that you can survive if you make a mistake with the precise moment. Because eventually, in one way or another, the final phase of each bubble always resolves itself. The only real question that remains is whether you'll see it coming or whether you'll be the last to know, like most people.
If this type of analysis helps you see the markets more clearly, be sure to subscribe, because this is exactly the pattern I'll continue to break down as this cycle continues to unfold. See you next time
Comments 0