저자: StanleyDruckenmiller · 외국인
요약 보기
AI 붐이 차입(채권)에 크게 의존하고 있으며, 대형 대출기관이 AI 관련 부채를 조용히 털어내는 징후가 있다고 주장. 순환 매출(A가 B에 투자 후 B가 A 제품 구매)과 주식-채권 스프레드 괴리를 근거로 2007년과 유사한 패턴이라 경고. 다만 기술 자체는 실재하며 예측 자체보다 경고와 점검을 강조. 구체적 가격 방향 예측은 명시되지 않음.
다듬은 글
Now, somewhere on a trading floor in New York, a bank is quietly backing away from the rate it publicly stated to you about safety. I've seen this same scenario before. I saw him in 2000. I saw it in 2007 , and I'm watching it happen again. Right now, in the midst of the AI boom , everyone tells you it's unstoppable .
I want you to remember this phrase. Banks are not panicking loudly. They panic quietly in memos you will never see while the television tells you everything is fine. This is exactly what happened until 2008. Big banks were getting rid of mortgage risks behind closed doors while their own analysts sat on CNBC telling ordinary people to keep buying, keep borrowing, keep believing that home prices were only going to go up.
A few weeks ago, the same pattern manifested itself again. This time, we owe it to artificial intelligence. If you have a 401k, pension, or retirement account that you never really paid attention to, it affects you, whether you chose it or not. No one will come into your living room and warn you. So stay with me because by the end you will understand exactly what is happening, why this is important for your money, and exactly what to do about it before everyone else does.
Where does the money for the artificial intelligence boom really come from? Let's start with something most people have never really thought about. Where does the money for the artificial intelligence boom really come from? You would think that a company of this size that makes headlines would only use its own profits, right?
They make billions every quarter. Wrong. Or at least not the whole story. A huge amount of the money funding data centers, chips, power plants, and artificial intelligence infrastructure is now borrowed. It is not accumulated over years, it is not paid in cash , which is on the balance sheet . Borrowing is similar to how a family borrows money for a house that they cannot afford with their current income.
Only this time it's hundreds of billions of dollars, and the house is a bet that artificial intelligence will eventually generate enough profit to pay it all back with interest. Think carefully about what this really means. These are n't small, sloppy startups that risk money that no one will miss. These are some of the largest and most trusted companies on Earth, entering the bond market and asking to borrow more, faster, than has ever been borrowed in any single industry in the history of finance.
Not railways, not point growth. Nothing can match this pace. So who is really on the other side of this transaction? Who lends them money? This is n't some mysterious billionaire sitting in a back room writing checks. These are bond funds. These are pension funds. These are pension funds that sit quietly in your 401k retirement account, purchasing what is automatically designated as safe investment debt, without you having to click a button, sign a form, or ask a single question.
Your money, without your permission, is very likely already somewhere in this bet. This should bother you a little , and it should bother you even more when you understand how this cycle actually works , because this is not a one-time transaction. This is a cycle. These companies borrow to build. Construction is called progress.
Progress is used to justify even bigger loans. And pension funds holding securities at the other end of this cycle continue to buy, because buying worked well for years until, historically, it suddenly stopped. rhymes with 2007. This is where it starts to rhyme with 2007. And I mean rhyming in a way that should really bother you.
Not in some vague abstract sense, but in the mechanics themselves. Back then, banks would write mortgages, package thousands of them together into bonds, and sell those bonds to the same pension funds I just mentioned. Mortgages financed even more mortgages. Bonds financed even more bonds. And everyone along the chain received a commission for their part.
So no one had any real incentive to stop the music, even people who privately suspected something was wrong. Then , quietly in the summer of 2007, the smart money on Wall Street began to back away from this deal, without announcing it publicly, without warning their retail clients, just quietly reducing their own exposure while continuing to tell the public that everything was fine.
The fundamentals were solid. It was a healthy market. I experienced this period professionally. I watched portfolios I respected, managed by smart, careful people, go bankrupt in a matter of months because the people who ran them found out what was really going on, after the insiders had already reoriented themselves.
There is now evidence of the same quiet reorientation in the AI-related debt markets. Financial statements have emerged that show big lenders are working behind the scenes to get rid of AI-related debt , repackaging it, spreading the risk among smaller buyers, trying to find new demand before the broader market fully understands how much exposure has accumulated at one time.
This is not something that appears as a flash of red headline on the evening news. This is what happens quietly in the background in research notes and internal memos while the public conversation remains optimistic, full of talk of unlimited future demand. I want you to think about this contrast for a second, because it corresponds to the most important pattern of your West.
I separate myself from doors and superficial agents, but with roso majes and circular income. Talk about something that should really concern anyone who understands how a real healthy business actually works. Circular income. It's an old trick, not a new one, but it reliably shows up right before the bubble bursts, almost like a fingerprint left at the scene.
Here's how it works, in plain language . Company A invests several billion dollars in Company B. Then Company B turns around and spends that money to buy products or cloud services from Company A. Then Company A reports the expenses as revenue growth on its earnings report, and the stock jumps because Wall Street loves growth.
On paper, this looks like huge organic customer demand. In fact, the money never left the building. They just walked in circles and returned home with another shortcut on the way out the door. This is also not a conspiracy theory. You can trace several such connections through public documents and financial statements right now, if you know where to look.
Some of the largest AI companies in the world have investment and purchasing relationships with each other that are almost perfectly self- centered . One company funds a partner, the partner turns around and buys computers or chips from the same company that funded it, and both parties quietly record this as growth on their respective books.
When independent analysts began looking for hard evidence that business truly exists outside the tech industry , regular companies in retail, manufacturing, and healthcare were actually making significantly more money thanks to the implementation of artificial intelligence. The evidence so far has been weak, much weaker than it would seem to be, judging by the assessments.
This should stop you for a moment. If the main companies profiting from AI right now are mostly companies selling AI to each other, that's not really a growth economy in the traditional sense. It's a vicious circle. A vicious circle that looks like growth from the outside , but in reality it's just the same dollars spinning faster and faster.
And closed cycles, no matter how complex or well- founded , eventually run out of fresh money to keep the circle going, because at some point someone in that cycle needs real outside money to keep paying the interest on all that borrowed money. What the bond market is quietly saying, I want to bring you back to something concrete now because numbers matter a lot more than opinions, including mine.
When I was a professional money manager, we watched bond spreads the way a doctor constantly, almost unconsciously, watches a patient's pulse. In simple terms, the spread is simply the extra interest a company has to pay compared to the safest possible borrower, usually the government. When spreads widen for a particular company or an entire sector, it means that lenders are nervous and are demanding more compensation to continue taking on that risk.
Recently, bonds linked to some of the biggest names pushing artificial intelligence infrastructure have expanded noticeably, even as the stock prices of those same companies were still celebrated nightly on financial television. This gap between what the stock market is cheering about and what the bond market is quietly factoring in behind the scenes is one of the most reliable warning signs I know of at this point.
Stock market investors are getting emotional. They chase headlines. They chase momentum. Bond investors do math, cold, unemotional math, because their downside is real and immediate. If a borrower can't pay when people who calculate based on math start to back off , while emotional people are still buying with both hands, it's no coincidence.
This is a signal, and it is usually early. I observed this exact divergence with mortgage-backed securities back in 2007. Construction company stocks continued to rise even as the underlying mortgage bonds beneath them were already cracking. No one wanted to hear about it then . And those who listened were called alarmists until they stopped being them.
How does this actually unfold? Now let's look at what would actually happen if this were to unfold, because I don't want to just scare you and leave you in a hopeless situation. I want you to really understand the mechanics step by step, as I would take a client through it. It all starts with creditors.
If bond buyers continue to demand higher yields to continue lending to these companies, borrowing becomes more expensive overall. And when loans become more expensive, spending plans are cut because these developments were never designed to operate at higher interest rates. We're talking trillions of dollars in planned data center construction, chip procurement, and energy infrastructure.
All of this is built on the assumption that cheap and easy loans will continue more or less indefinitely. At a time when loans are becoming even more expensive, some of these plans are being postponed, scaled back, or quietly shelved altogether. When spending is cut in this way, the cyclical revenue I just described begins to quickly dry up.
Because remember, a significant portion of it never came from external customers in the first lease period. It came from these companies, which bought from each other in this closed loop. When their income disappears or even just noticeably slows, profits are disappointing. And when earnings disappoint at companies that together make up nearly a third of the entire stock market, the market doesn't just sag a little around the edges, it drags almost everything else down with it.
Index funds, the ones that are in most people's retirement accounts right now, are forced by their nature to hold these same companies in huge chunks, whether the fund manager wants to or not, because that's how index expectations work. It's domino one, domino 2, domino 3, domino 4, and the fifth domino has your name written on it, it's quietly sitting in your retirement account.
Because that account almost certainly holds both stocks directly and, indirectly through various bond funds, a certain significant portion of that debt as well, even if you never looked closely enough to notice it. the gap between public trust and private positioning. Let me give you another piece of this puzzle, because it directly connects everything I just said to how public sentiment is managed.
Whenever there is a quiet reduction in risk on the trading desks, it is usually followed by a wave of loud, optimistic public research that predicts even more growth in the future. This does not necessarily indicate coordination. Maybe it's just the natural course of a large, complex financial system.
But the pattern itself is worth knowing, as it means that good headlines and confident forecasts do not automatically mean that the underlying risk has disappeared. Sometimes they appear at the very moment when insiders are already quietly reducing their own influence. Observing this gap between public trust and private positioning is one of the most useful habits I have ever developed in this business.
And this is the one I would advise you to start practicing on your own. Not just with artificial intelligence, but with any investment story that suddenly seems too good to be true . To be fair , the technology is real. I have to be fair because I don't believe in scaring people without giving them the full honest picture.
That's not what I'm talking about. The technology itself is real. And I want to be very clear on this issue. I'm not sitting here telling you that artificial intelligence is fake, worthless, or some elaborate hoax . That would be foolish, and frankly, it would insult your intelligence and mine . The chips really work.
The software is truly improving month after month. Truly useful products are already being created that are used by millions of people. I'm not questioning the technology itself. It's the funding structure behind it, and more specifically, the growing gap between how much has been borrowed against future promises and how much real, off-cycle, sustainable revenue currently exists to actually support all of those loans.
Bubbles historically almost never form around fake technologies . This is an important distinction that people overlook. The railways were absolutely real and truly changed the country. The internet was absolutely real and truly changed the way we all live and work. Housing, people really needed houses to live in.
This demand was never false. The reality of the underlying technology or asset has never really determined whether a bubble will eventually burst. Each time, it determines whether the funding has significantly outperformed the fundamentals supporting it. Right now, by almost every metric I personally track, AI funding is sharply outpacing proven external sustainable demand.
This gap must eventually narrow, one way or another. There is no option where it will simply remain open forever. It either shrinks because real income from real external customers eventually catches up over time, which would be a good outcome for everyone involved, or it shrinks because cheap loans simply stop and the whole structure has to snap back to reality all at once, which is an outcome that really hurts ordinary people who never anticipated this and had no way to prepare.
What you should actually do. So, here's what I really want you to do with all of this. Because pointing out a problem without giving you a plan is not worth your time, and I take that seriously. First, find out what is actually in your retirement account today, this week, not someday. Most people have never really watched.
Open your 401k or retirement fund statement and see how concentrated it is in a small number of large technology companies. Because if it's built around a standard index fund, which most standard retirement plans are, the answer is probably a lot more concentrated than you might think right now. Second, understand that diversification is not just a buzzword your financial advisor uses to sound smart.
This is an actual practical tool that protects you when one particular sector is heavily overloaded compared to all the others. If almost a third of your money is actually tied up in the fortunes of six or seven companies, you don't really have a diversified portfolio in any meaningful sense. You have a highly concentrated bet that you never consciously made.
Third, and this is the step people most often skip, have a specific written plan for what you will do if the sector adjusts sharply and suddenly , rather than a vague feeling that I will figure it out when it happens. An actual plan with specific steps. Because when real volatility hits a portfolio, people who have already decided in advance what they will do tend to stay calm and act rationally, while people who have never decided anything in advance make their worst financial pause decisions right in the middle of the worst possible moment.
Almost always selling closer to the bottom due to pure understandable panic. This one simple difference – having a plan before you need it, not after – is the difference I have personally observed that separates investors who have emerged from past downturns relatively unscathed from those who were truly destroyed and took years to recover.
This pattern repeats itself. I've been doing this for a long time, long enough to watch this same basic story repeat itself over and over again in different costumes. Surface details change every time. Mortgages are turning into AI bonds . Stocks are becoming chipmakers. But the underlying human model that drives this never really changes, because human behavior itself doesn't change much from decade to decade.
People see that something is increasing in price. They naturally assume that it will just continue to grow. Institutions begin to lend more aggressively, contrary to this comfortable assumption, and lending becomes increasingly aggressive the longer nothing noticeably bad happens. until the moment when something finally happens.
And it usually happens faster than anyone expected. I'm not telling you all this because I like being the person who delivers bad news. Believe me, no. I would sincerely prefer to be wrong about all of this and just watch these AI companies quietly and steadily grow to their current valuations over the next decade.
That would be the best outcome, for sure. But I learned from my own bitter experience over several cycles that hoping for a mistake is never a strategy. As is the tacit assumption that someone else, some regulator, some bank, some fund manager sitting in an office, will shape your particular money more carefully than you do.
No one does this for you. This is not cynicism. It's just something that 30 years in this business have taught me over and over again, through my own bitter experience. So here's my direct challenge to you before you wrap this up and move on with your day. Don't just treat it as passive entertainment and forget about it until tomorrow morning.
Check your retirement account this week. Not eventually, but this week. Take a close look at what exactly you are holding inside. If you don't fully understand what you 're looking at, it's not such a minor problem. That's really the whole point of the problem, because you really can't protect yourself from a risk that you can't even clearly see in front of you.
This is not a prediction of imminent doom. I want to clarify this again. This is a warning meant to help you prepare, and these are two very different things. Only one of them actually protects you.
Comments 0