저자: AdamFeather · 외국인
요약 보기
결론: 영상 발화자는 마이클 버리가 AI 칩 시장에서 1967~1970년 컴퓨터 리스 버블과 같은 거품이 재현되고 있다고 본다는 점을 소개하며, 과거 사례를 근거로 경계를 권하지만 구체적인 종목 매수·매도 지시는 없다. [배경 및 사례 요약] 영상은 1968년 출간된 『The Money Game』(George Goodman, 필명 Adam Smith)의 1967년 장면을 재구성한다. 3명의 젊은 트레이더가 컴퓨터 임대 회사 주식에 대부분 차입으로 투자해 수익을 냈고, 1967년 한 트레이더는 5,000달러를 6개월 만에 50만 달러 이상으로 늘렸다. 그는 자기 자금 5%만 투입했다고 설명된다. [리스 사업 구조] IBM 시스템 360은 임대료를 약 5년 감가상각 가정으로 책정했으나, 리스 업체들은 같은 기계를 최소 10% 저렴하게 임대하며 약 10년 사용을 가정했다. 감가상각 기간을 길게 잡아 장부상 이익이 크게 늘었고, 이 가정으로 대규모 차입 매입이 이뤄졌다. 타임지는 1967년 이 위험을 경고했다. [결과] 1970년 5월 평균 컴퓨터 주식은 약 80% 하락했다. 기계 수명은 가정과 비슷했으나(Stephen Oliner 분석 기준 시스템 360은 약 8년 9개월 사용), 1970년 6월 IBM 시스템 370 출시 후 구형 기계의 임대 가격이 급락했다. 메인프레임의 품질 조정 가격은 10년 넘게 연 약 20%씩 하락했다. 1968년 고점에서 미국 주식에 투자한 사람은 인플레이션 조정 후 1990년대에야 원금을 회복했다고 언급된다. [당시 시장 분위기] Gerald Tsai, Bernie Cornfeld(IOS 붕괴), 상호 명칭에 'tronics'를 붙여 주가가 뛴 사례, 2017년 Long Island Iced Tea의 사명 변경(Long Blockchain) 후 주가 급등, 1970년대 Nifty Fifty 종목(폴라로이드 등)의 급락이 소개된다. [현재와의 비교 주장] 발화자는 엔비디아를 IBM, AI 칩을 시스템 360에 비유하고, CoreWeave 등 신생 전문 업체의 데이터센터 임대 사업을 과거 리스 업체와 대응시킨다. 칩 수명이 5~6년 정액 감가상각되는 점을 지적하고, 엔비디아가 자사 발표 자료에서 첫해 40% 이상 가치 하락을 보여줬다는 점을 근거로 든다. 또한 엔비디아가 수백 개 AI 기업에 지분을 보유한 점을 순환 구조 위험으로 언급한다. [출연자 전망] 마이클 버리는 AI 칩 수요는 실재하지만 감가상각과 세대교체로 인해 이익의 지속성에 의문이 있으며, 시장이 과열 상태라고 본다는 취지다. 그는 2025년 말 자신의 펀드를 외부 투자자에게 닫았고, 자신이 보는 주식 적정 가치가 시장과 한동안 맞지 않았다고 서한에 밝혔다. [실천 조언(영상 내용)] - 보유 자산이 두 배가 되면 원금만큼 매도해 원금을 회수하고 나머지는 시장 돈으로 본다. - 매도 자금은 대출 상환이나 현금으로 보관한다. - 주가 급등 시 '쉽게 돈 번' 사람의 과도한 차입, 사명 변경, 증권 팁 확산을 과열 신호로 본다. - 버핏(1974년 현금 보유)처럼 폭락 시 저가 매수 여력을 확보한다. [리스크 및 한계] 영상은 과거 사례를 현재에 대입하는 비유 중심이며, AI 칩 수요 지속 기간이나 엔비디아 실적에 대한 구체적 수치 전망은 제시하지 않는다. 특정 종목의 목표가나 매매 타이밍은 언급되지 않았다. 영상 말미에 무관한 광고성 문구(퀴즈, 맞춤 조언 안내)가 섞여 있다.
다듬은 글
New York, late 1967. A famous speculator sits in his office with three young traders he calls kids, none older than 29. They never look up from his ticker tape. One of them has turned $5,000 into more than half a million in 6 months. He did it with shares in companies that rented out computers, and he did it almost entirely with borrowed money.
On a cautious day, he put up 5% of his own cash. Asked to explain himself, he says that the world's need for computers is practically infinite. And that the profits of these companies will double every year for years and years to come. He was right about the computers. The world did need them endlessly.
The computers lasted almost exactly as long as the companies hoped. And by the spring of 1970, shares in those companies had lost roughly 4/5 of their value. , has just retold this story. He thinks it's happening again, almost line for line, with the chips behind artificial intelligence. And here is how bad it can get.
If you had bought the American stock market at its peak in 1968, then once you allow for inflation, you would not have got your money back for good until the 1990s. Burry is often early, sometimes by years. So, we will step into that office in 1967, learn the game the kids were playing, watch how it ended, and then hold it up against today.
And at the end, the one kid who kept his fortune and what he did differently. Part one, the kids' market. The scene comes from a book called The Money Game, published back in 1968 under the pen name Adam Smith. It was written by a financial journalist called George Goodman, who borrowed the name of the famous economist, presumably on the grounds that the original wasn't using it anymore since he was dead.
It became a best-seller. Warren Buffett recommended it to his own investors that summer, calling it full of insight and wit. Goodman went on to present a long-running money program on American public television under the same borrowed name. In the book, Goodman visits an old friend, a celebrated speculator he calls the great Winfield.
Winfield greets him with mockery. Goodman, he says, is the sort of person who asks how fast these computers are written off. You cannot make money asking that sort of thing, Winfield tells him. It shows you're middle-aged. Winfield's answer to the market is to hire children, not literally, but nearly.
He's taken on three young traders, gives each of them a small stake, and splits the profits. It's their market, he says. He's just along for the ride. Billy the Kid, the one with the half a million, is in computer leasing shares with names like Leaseco, Data Processing, and Randolph Computer. Goodman notices that they all seem to have the word computer or data processing somewhere in the name.
That alone in 1967 was worth something. Pause on that 5% because it matters at the end. Putting up so little of his own money meant a 5% rise doubled it, and a 5% fall would wipe him out completely. In a rising market, it feels like genius. It's really just a very short fuse. If 5% sounds reckless, consider that it is exactly the deposit on a 95% mortgage.
I have watched plenty of people stretch to buy houses they could only just afford on just that kind of borrowing, absolutely certain that prices could only go up. Billy the Kid would have recognized them straight away. Another of the kids is in oil shale. He explains that the western United States is sitting on more oil than the rest of the world put together, locked in rock, and that the technology to get it out is coming quickly.
A share at $24, he reckons, could one day earn $750 a share. And then Winfield asks the question himself before Goodman can. What makes a finance company worth 50 times its earnings? The kid smiles politely at the old generation who can't be expected to understand the new economics, and the scene moves on.
Burry's point in retelling this story is that this was published almost 60 years ago, and every line of it could have been written this morning because of the fact that human nature doesn't really change. Swap computers for AI chips, swap oil shale for whatever it is fashionable this week, and you wouldn't need to change much else.
Human nature, as we've said, is more constant than almost anything in markets, which brings us to part two, the leasing game. So, what were these computer leasing companies actually doing? Well, in the mid-1960s, IBM dominated computing. Its great computer was the System 360, a room-sized mainframe that companies used for payroll, stock control, and accounts.
IBM had staked almost everything on it. Most firms didn't buy one, they rented it directly from IBM. Developing the 360 cost about $5 billion in 1960s money, more than the company's whole yearly revenue, and the press at the time called it IBM's $5 billion gamble. That bet paid off spectacularly. The leasing companies spotted a gap.
IBM set its rents on the assumption that each computer would pay for itself over about 5 years. The leasing companies bought the same machines from IBM, then rented them out at least 10% cheaper, betting the computers would stay in use for around 10 years. Spread the cost over twice as long, and you can charge less and still make a fortune.
Time magazine wrote about it in 1967. The customers were the biggest names in America. General Motors, Boeing, AT&T, Mobil, and curiously IBM didn't mind at all. The leasing companies were buying its machines with cash, and IBM needed vast amounts of cash to pay for its next generation. The profits looked wonderful.
In the accounts, a computer written off over 10 years cost half as much each year as when written off over five. So, the leasing companies reported huge earnings, their shares soared, and they borrowed more to buy more machines from IBM. One of them was run by a young man called Saul Steinberg, who had started his company with $25,000 borrowed from his father.
He was willing to write his computers off over eight years rather than four. That one assumption made him rich. Even at the time, Time magazine flagged the danger. The leasing company's profits could plunge, it warned, leaving them with a mountain of debt if the next generation of computers arrived sooner than expected.
Which brings us to part three. Write about everything except the price. Years later, an economist at the Federal Reserve, Stephen Oliner, went back and worked out how long IBM's System 360 machines had actually stayed in service. The answer was eight years and nine months. The leasing companies had been almost exactly right about that.
Even the oil shale kid turned out to be right in a sense. America did eventually unlock enormous amounts of oil from rock in the fracking boom of the 2010s. He was only about 45 years early. Anyone who borrowed to back him in 1967 was long broke by the time he was proven correct. And the appetite for computing was every bit as big as the kids had said.
It still is. It made no difference. Computer stocks peaked in late 1968. By May 1970, the average one had fallen about 80. What went wrong was not how long the machines lasted for. It was what they could be rented for once something better appeared. In June 1970, IBM launched its next generation, the System 370.
Every customer now wanted the new one. The old machines were still working perfectly, but nobody would pay much to rent them. Picture those offices in May 1970. Paneled walls, the ticker tape still chattering, but every price on it lower than the one before. Lenders on the telephone, and the computers they own still humming away in customers' basements across America, earning a little bit less every month.
Bowie's favorite detail is the timing. The leasing shares hit their lowest point in May 1970. IBM's new computer was launched 5 weeks later. The market didn't wait to see the new computer before it gave up on the old business. And after that, things only got worse for the old machines, not better. IBM's computers kept getting cheaper and better.
The useful life of each generation shrank from nearly 9 years for the 360 to about 6 for the 370 and 5 and 1/2 for the one after that. Quality-adjusted prices for mainframes fell about 20% a year for more than a decade. The leasing companies were stuck with fleets of machines rented at prices they could no longer get and debts they had taken on when the future looked endless.
Through the 1970s, most of them collapsed or were swallowed up. Warren Buffett had seen enough even earlier. In 1969, with speculation everywhere, he wrote to his partners that he was closing his investment partnership because he had run out of bargains he understood. He was 38 and to the young men making fortunes on the ticker tape, he looked like a relic.
Five years later, in 1974, with shares down and nobody interested, a reporter asked him how he felt about the stock market. He said he felt like an oversexed man in a harem and that this was the time to start investing. That is the other half of the story because he had stepped away from the frenzy, he had cash when everything was cheap, and around at the same time Buffett was quietly buying shares in the Washington Post at a fraction of what the business was actually worth, and he held onto those shares for decades.
There is a modern echo [clears throat] of Buffett's exit. Late last year, Michael Burry closed his own fund to outside investors. In his letter, he said his estimate of what shares were worth had not been in step with the market for some time. Like Buffett in 1969, he didn't wait for the proof. And the 1970s were not kind to anybody holding shares.
There was an oil crisis in 1973, petrol queues, prices doubling over the decade, and two brutal falls in the stock market. Flares and sideburns were the least of it. The wider market didn't recover quickly, either. 14 years after that peak in August 1982, American shares were still below it, even before counting a decade of savage inflation.
Hence, the long wait. Think about what that means in a life. Someone who had put their savings into American shares at the top in 1968, age 35, would have been close to 60 before those savings were worth as much again in real terms. Which brings us to part four, the go-go years. The kids were not alone.
The late 1960s were called the go-go years, and they produced characters nobody would believe in a film. There was Gerald Tsai, the most famous fund manager in America. In 1966, he launched a new fund and hoped to raise $25 million. People queued to give him 247 million. Within 3 years, his fund was one of the worst performers in the country.
But Tsai himself had already sold his management company in 1968 for around $30 million. He had cashed in before the dream ran out for everybody else. Then there was Bernie Cornfeld, who ran an enormous fund empire called Investors Overseas Services from a chateau near Geneva, surrounded by fast cars and film actresses.
his salesmen were recruited with a single question, "Do you sincerely want to be rich?" Thousands did. In 1970, the empire collapsed. The man who took it over, Robert Vesco, was accused of making off with over $200 million and spent the rest of his life on the run, ending up in Cuba. And there were the names.
The economist Burton Malkiel later recalled how in the 1960s, companies discovered that adding the word tronics to their name was worth more than any product. The maker of musical instruments rebranded itself Astro tronics. The share price, naturally, took off. If you think we are too sophisticated for all of that now, consider the winter of 2017.
A company called Long Island Ice Tea, which made iced tea, changed its name to Long Blockchain. Its shares nearly tripled in a day. It did not, as far as anybody could tell, change the tea. Then, a few years after the leasing bust, came the Nifty Fifty. Fifty great American companies that were supposed to be so reliable you could buy them once and never sell.
Polaroid was one. In 1972, investors paid about 90 times its earnings. Two years later, its shares had fallen about 90%. The cameras were still excellent. Every one of these stories has the same shape. A real idea, a great story, a crowd of buyers too young or too excited to remember the last time, and a price that assumed nothing would ever go wrong.
Which brings us to part five, the rhyme. So, why does Burry think this is all happening again? Put Nvidia where IBM was, and its AI chips where the System 360 was. In place of the leasing companies, put the businesses building data centers and renting out computing power from the tech giants themselves to newer specialist firms such as CoreWeave.
Burry even set Billy the Kid's 1967 sales pitch beside today's market and invited his readers to squint. "Infinite demand, earnings doubling every year, the surface barely scratched." You could put those words into the mouth of almost any AI enthusiast on television this week, and nobody would notice that they were nearly 60 years old.
Today's version of Tronix is two letters. Listen to any company results this year and count how many times you hear AI. It has become the word you add to a slide to make a share price stand up just a little bit straighter. The demand is real, just as it was in 1967. The machines will probably last a reasonable time, but Nvidia now releases a new generation of chips every year, and each one makes the last one cheaper to replace.
Its next family of chips, Nvidia itself claims, will produce artificial intelligence at a fraction of the cost of the current one. Meanwhile, the companies buying the chips write them off slowly in equal slices over five or six years. Burry notes that not one of the big buyers uses a faster schedule, even though Microsoft's own finance chief has described these chips as short-lived assets.
Write something off slowly and your profits look higher today. That was Sol Steinberg's trick in the 1960s. Nvidia has even handed Michael Burry an argument. It recently showed investors a slide claiming that its chips hold their value. On the same slide, it drew its own idea of how fast they should be written off.
The steep drop, losing over 40% of their value in the first year alone, far faster than its own customers assumed. Burry called that a gift to people like him. He also points out a strange echo in the money. In 1967, Time found it odd that IBM welcomed the leasing companies, since they paid it cash it badly needed.
Today, Nvidia holds stakes in hundreds of private companies across the AI industry, many of them are buyers of its chips. And the question that Winfield asked back in 1967 is precisely the one that Michael Bury asks right now. Somebody has to rent these chips out at prices that pay for them, while the next generation makes them look expensive.
So, what makes that business worth 50 times its earnings? Which brings us to part six, the one who got out. In the book, Winfield knows exactly how all of this will end. "All of these kids but one," he says, "will be broke. And that one will be a multimillionaire." The one was Saul Steinberg, the young man who had begun with his father's loan.
Steinberg saw where the game was heading. While his shares were still soaring in 1968, he used them to buy something real, an old Philadelphia insurance company called Reliance, which had a great deal of cash. He swapped paper that was about to collapse for a business with money in the bank. Congress was so alarmed by deals like his that it held hearings, and his team had used a code name for the takeover so that even their own staff would not know the target.
It made him one of the richest men in America for decades. He ended up living in John D. Rockefeller Jr.'s old apartment on Park Avenue, 34 rooms. For his 50th birthday, his wife threw a party at which live actors stood motionless in giant gilt frames, recreating the old master paintings from his collection.
If you were looking for the moment the 1980s peaked, that is a strong candidate. He became one of the most feared corporate raiders of the decade. In 1984, he went after Walt Disney, and Disney paid him a very large sum to go away, which is not something Mickey Mouse does for just anybody. What set him apart was not that he was cleverer about computers than the others.
They all understood computers. He alone saw his soaring share price as something to spend before it vanished, and he spent it on an old dull business with real cash, which was insurance. So, this brings us to part seven, what you can do. You probably don't own a computer leasing company yourself. But, if you have a pension or an index fund, you own a slice of the modern version of this story whether you meant to or not.
So, here is how to spot when you're standing in Winfield's office and what to do about it. The legend goes that in the summer of 1929, Joseph [snorts] Kennedy, father of the future president, was having his shoes polished on Wall Street when the shoeshine boy gave him a share tip. Kennedy reasoned that if the shoeshine boy was in the market, there was nobody left to buy, and he sold.
A few months later came the crash. Whether or not it happened quite like that, the test still works. When the people giving you share tips have never owned a share before, the party is late. A second sign is the gastropub. When a company adds a fashionable word to its name and the share price jumps on the word alone, it's the same as a pub that puts gastro on the sign and doubles the price of the chips.
The chips have not changed. A third sign is the 5% deposit. When people you know are borrowing heavily in order to buy something because it can only go up, whether that's shares, crypto, or a house at the edge of what they can afford, Billy the Kid is back at the ticker tape. And the last sign is you.
If something you own has done so well that you would feel a fool selling any of it in case it doubles again, that is exactly how Billy felt in 1968. So, what do you actually do? Here, the advice is simple, and it comes from Saul Steinberg himself. When something you own doubles, sell enough of it to get your original money back.
That's it. What's left is the market's money, and you can let it ride with a clear conscience. If it keeps on going up, you're still in. If it collapses, you've lost nothing that you started with. Then, put what you took off the table somewhere dull. Pay down the mortgage, build a cash pile. Buffett had cash in 1974, which is why he could go shopping while everybody else was trying to sell.
Burry's own advice to most people is just as modest. Sell some of what has done well and keep some cash for the bargains every crash leaves behind. It will feel like cowardice at the time. Saul Steinberg would call it the whole point. Because the most unsettling thing about this entire story is how ordinary a bubble feels from the inside.
Nobody in Winfield's believed they were in one. They thought they were early. They had a good story, real demand, and the numbers mostly right. They just forgot that sooner or later somebody builds a better computer. In 1967, the kids were right about everything except the one thing that actually mattered.
The computers were needed and they did last and the shares still fell 4/5 5 weeks before the proof that that would happen finally arrived. Michael Burry thinks the kids are back with different machines and the same certainty. He may be early, but as Adam Smith wrote back in 1968, we have all been here before.
his own accounts rather than the right nd covering similar topics as well as the free quiz which helps you evaluate your own position and gives you tailored personalized advice to move you forward. .
Comments 0