Leverage Kills. Patience Pays. A Look at Aschenbrenner, Buffett, and Jhunjhunwala
Rahul Chavan | July 31st 2026
There is an old rule in investing. You can be smart, you can be right, and you can still go broke. The one thing that ties every blowup together is leverage. The one thing that ties every great fortune together is patience. This week gave us a fresh, painful example of the first rule. It also gave us a good reason to revisit the second.
What happened to Leopold Aschenbrenner's fund
Leopold Aschenbrenner is a young former OpenAI researcher. He became famous for a long essay called "Situational Awareness," which argued that artificial general intelligence was coming faster than most people believed. He turned that thesis into a hedge fund, backed by well known tech investors like Daniel Gross, Patrick and John Collison, and Nat Friedman. He bet big on AI. The bet worked, for a while. His fund reportedly grew from around $225 million to roughly $45 billion, riding a huge run in AI-linked stocks like semiconductor makers and cloud providers.
Here is the part that matters. He was not just betting on AI. He was betting on AI with borrowed money, reportedly around four times his own capital. That means for every dollar he actually had, he was controlling about four dollars of stock. When your bet goes your way, leverage feels like a superpower. Your gains get multiplied. When your bet goes against you, even a little, leverage turns a bruise into a bloodbath.
In late July 2026, AI-linked stocks like SK Hynix and CoreWeave dropped somewhere between 35 and 47 percent. That is a bad month, not a bad decade. Plenty of great investors have sat through drops like that and come out fine. But Aschenbrenner did not have that luxury, because his positions were pledged as collateral against loans from his prime brokers, Goldman Sachs, JPMorgan, and Bank of America. When the value of that collateral fell below what the loans required, the brokers issued margin calls. He needed to post more cash or more securities, fast, or the brokers would sell his positions for him.
He did not have the cash. So on July 30, 2026, his fund sold its entire public stock book, longs and shorts together, in a single block trade before the market opened. The buyer was Ken Griffin's Citadel, which picked up the portfolio at a discount, because that is what happens when you are a forced seller. Aschenbrenner's fund reportedly shrank from around $45 billion to about $10 billion in a matter of days. The one thing he kept was a private stake worth about $5 billion in Anthropic, which was not part of the public, leveraged book.
To be clear about what actually happened: his research was not necessarily wrong. AI stocks had gone up over 1,000 percent under his watch before the crash. The problem was not his judgment about the future. The problem was that he built a position so leveraged that he could not survive being wrong, or even survive a normal, temporary dip, long enough to be proven right.
The same story, explained like you are five
Imagine you have ten dollars. You really believe your friend's lemonade stand is going to be the best one on the block. So you buy ten dollars of lemonade stand tickets.
Now imagine a grown-up offers to lend you thirty more dollars, so you can buy forty dollars of tickets instead of just ten. If lemonade sales go up, you make four times as much money as if you had only used your own ten dollars. Sounds great.
But the grown-up says one thing. "If the value of your tickets ever drops too much, you have to pay me back right away, even if that means selling your tickets when the price is low." One rainy afternoon, fewer people want lemonade, and the price of the tickets drops a little. Not a lot. Just a little. But because you borrowed so much, that little drop is a big problem for you. The grown-up says pay up now. You do not have thirty extra dollars sitting around, so you have to sell all your tickets right away, at a bad price, to someone else who is happy to buy them cheap.
If you had only used your own ten dollars, that rainy afternoon would not have bothered you at all. You would have just waited for sunny days to come back.
That is the whole lesson. Borrowed money makes good days feel great and bad days force you to sell exactly when you should be holding on.
Why long term investing works
Warren Buffett has said that the stock market is a device for transferring money from the impatient to the patient. Good businesses grow in value over years and decades, not days and weeks. But their stock prices bounce around every single day for reasons that have nothing to do with the business itself: interest rate news, a bad headline, a scared seller, general mood swings.
If you own a good business without debt hanging over your head, those bounces do not threaten you. You can simply wait. Time does the heavy lifting. A wonderful business compounds its earnings year after year, and if you hold on long enough, the stock price eventually reflects that. Mohnish Pabrai, who studied Buffett closely, likes to say investing is about a few big decisions, not constant activity. Find a good business, buy it at a fair price, and then do almost nothing. Let boredom be your strategy.
Long term investing also lets compounding work in your favor. A dollar that grows at a steady rate for thirty years turns into far more than most people intuitively expect, simply because growth builds on top of growth. But this only works if you are never forced to sell. Selling early, especially selling at a low point because you have to, is the single biggest way people destroy their own returns.
Why leverage is the enemy of patience
Leverage and patience cannot coexist for long. The whole point of long term investing is that you get to choose when to sell. Leverage takes that choice away from you. A lender or a broker gets to decide, based on a formula, not based on whether your reasoning about the business was right.
This is why Aschenbrenner's story is not really a story about AI stocks being a bad bet. It is a story about the danger of being right on the idea but wrong on the structure. Even Buffett, who has been enormously right about American business for seventy years, has almost never used leverage in a way that could force him to sell. He keeps huge piles of cash on hand specifically so that no one can ever make him sell a good business at a bad time. Charlie Munger put it simply: to get rich, you do not need leverage, you just need to avoid being wiped out. A small chance of a total wipeout, multiplied across a long enough career, eventually catches up with almost everyone who uses heavy leverage. That is the whole moral of the Aschenbrenner story, and it is not new. It happened to Long-Term Capital Management in 1998, it happened to countless investors in 2008, and it happened here in 2026.
Warren Buffett's big bets
There is a well known idea, often traced back to Charlie Munger, that Berkshire Hathaway's extraordinary results did not come from thousands of small, clever trades. They came from a small number of very large, very concentrated decisions, held for a very long time. Munger once said Berkshire's record was built on something like ten or twelve truly good decisions over fifty-plus years, not constant trading.
Look at the current numbers and the pattern holds up. As of early 2026, Berkshire's top five public stock holdings, Apple, American Express, Coca-Cola, Bank of America, and Chevron, made up about 67 percent of the entire public equity portfolio. Apple alone was worth more than a fifth of it. And that is before counting Berkshire's wholly owned businesses like GEICO, BNSF Railway, and See's Candies, which are not even in that public stock count but which have generated a staggering amount of the actual wealth over the decades. See's Candies, bought for $25 million, has thrown off more than $2 billion in profit over time.
The exact number, whether it is four bets or ten or twelve, is less important than the pattern. Buffett did not get rich by owning a little bit of everything. He got rich by finding a small handful of wonderful businesses, buying them without borrowed money, and then simply refusing to sell for decades. That patience, not any single stock pick, is the real skill.
Rakesh Jhunjhunwala's Titan bet
Rakesh Jhunjhunwala, often called India's Warren Buffett, built much of his legend on one stock: Titan Company, the watches and jewelry maker owned by the Tata Group. He started buying Titan around 2002 and 2003, at roughly 30 to 35 rupees a share, when almost no one else was paying attention to it.
Over the next two decades, Titan turned into one of the greatest compounding stories in Indian markets, rising well over 10,000 percent from his entry price. By the time of his death in 2022, Titan was the single largest holding in his portfolio, reportedly making up close to one-third of its total value. That is an enormous concentration for one stock to hold in a diversified portfolio, and it is a big reason his overall fortune, estimated at around 5 to 6 billion dollars near the end of his life, grew as much as it did.
It is worth being precise here, since the numbers get repeated loosely online. Titan was not literally "most" of his $5 billion portfolio in the sense of being over half. It was close to a third, which still made it by far his single biggest position and his single biggest source of gains. The rest of his fortune came from a long list of other holdings, companies like Crisil, Escorts Kubota, Star Health, and several others, plus new ventures he backed directly. But if you had to point to one decision that defined Jhunjhunwala as an investor, it would be Titan: spot a quality business early, take a real position in it, and then simply hold it through twenty years of ups and downs without flinching.
Putting it all together
Three stories, one lesson. Aschenbrenner had a good idea, wrapped it in heavy leverage, and lost most of the value in days when the market dipped and his lenders called. Buffett had good ideas too, wrapped them in patience and zero forced-selling risk, and built one of the greatest fortunes in history over seventy years. Jhunjhunwala did the same thing in India with Titan, buying early, sizing it big, and refusing to sell for two decades.
The difference between these outcomes was never really about who was smarter or who understood the future better. It was about who could afford to be patient. The investors who avoid borrowed money give themselves the right to wait out bad weeks, bad months, even bad years. The investors who lean on leverage hand that right away to someone else, a bank, a broker, a lender, who does not care about your thesis and will sell you out the moment the numbers say so.
If there is one thing worth taking from this week's news, it is this. Find good businesses. Buy them with money you actually have. Then be boring enough, and patient enough, to just hold on.
Leverage Kills. Patience Pays. A Look at Aschenbrenner, Buffett, and Jhunjhunwala
Rahul Chavan | July 31st 2026
There is an old rule in investing. You can be smart, you can be right, and you can still go broke. The one thing that ties every blowup together is leverage. The one thing that ties every great fortune together is patience. This week gave us a fresh, painful example of the first rule. It also gave us a good reason to revisit the second.
What happened to Leopold Aschenbrenner's fund
Leopold Aschenbrenner is a young former OpenAI researcher. He became famous for a long essay called "Situational Awareness," which argued that artificial general intelligence was coming faster than most people believed. He turned that thesis into a hedge fund, backed by well known tech investors like Daniel Gross, Patrick and John Collison, and Nat Friedman. He bet big on AI. The bet worked, for a while. His fund reportedly grew from around $225 million to roughly $45 billion, riding a huge run in AI-linked stocks like semiconductor makers and cloud providers.
Here is the part that matters. He was not just betting on AI. He was betting on AI with borrowed money, reportedly around four times his own capital. That means for every dollar he actually had, he was controlling about four dollars of stock. When your bet goes your way, leverage feels like a superpower. Your gains get multiplied. When your bet goes against you, even a little, leverage turns a bruise into a bloodbath.
In late July 2026, AI-linked stocks like SK Hynix and CoreWeave dropped somewhere between 35 and 47 percent. That is a bad month, not a bad decade. Plenty of great investors have sat through drops like that and come out fine. But Aschenbrenner did not have that luxury, because his positions were pledged as collateral against loans from his prime brokers, Goldman Sachs, JPMorgan, and Bank of America. When the value of that collateral fell below what the loans required, the brokers issued margin calls. He needed to post more cash or more securities, fast, or the brokers would sell his positions for him.
He did not have the cash. So on July 30, 2026, his fund sold its entire public stock book, longs and shorts together, in a single block trade before the market opened. The buyer was Ken Griffin's Citadel, which picked up the portfolio at a discount, because that is what happens when you are a forced seller. Aschenbrenner's fund reportedly shrank from around $45 billion to about $10 billion in a matter of days. The one thing he kept was a private stake worth about $5 billion in Anthropic, which was not part of the public, leveraged book.
To be clear about what actually happened: his research was not necessarily wrong. AI stocks had gone up over 1,000 percent under his watch before the crash. The problem was not his judgment about the future. The problem was that he built a position so leveraged that he could not survive being wrong, or even survive a normal, temporary dip, long enough to be proven right.
The same story, explained like you are five
Imagine you have ten dollars. You really believe your friend's lemonade stand is going to be the best one on the block. So you buy ten dollars of lemonade stand tickets.
Now imagine a grown-up offers to lend you thirty more dollars, so you can buy forty dollars of tickets instead of just ten. If lemonade sales go up, you make four times as much money as if you had only used your own ten dollars. Sounds great.
But the grown-up says one thing. "If the value of your tickets ever drops too much, you have to pay me back right away, even if that means selling your tickets when the price is low." One rainy afternoon, fewer people want lemonade, and the price of the tickets drops a little. Not a lot. Just a little. But because you borrowed so much, that little drop is a big problem for you. The grown-up says pay up now. You do not have thirty extra dollars sitting around, so you have to sell all your tickets right away, at a bad price, to someone else who is happy to buy them cheap.
If you had only used your own ten dollars, that rainy afternoon would not have bothered you at all. You would have just waited for sunny days to come back.
That is the whole lesson. Borrowed money makes good days feel great and bad days force you to sell exactly when you should be holding on.
Why long term investing works
Warren Buffett has said that the stock market is a device for transferring money from the impatient to the patient. Good businesses grow in value over years and decades, not days and weeks. But their stock prices bounce around every single day for reasons that have nothing to do with the business itself: interest rate news, a bad headline, a scared seller, general mood swings.
If you own a good business without debt hanging over your head, those bounces do not threaten you. You can simply wait. Time does the heavy lifting. A wonderful business compounds its earnings year after year, and if you hold on long enough, the stock price eventually reflects that. Mohnish Pabrai, who studied Buffett closely, likes to say investing is about a few big decisions, not constant activity. Find a good business, buy it at a fair price, and then do almost nothing. Let boredom be your strategy.
Long term investing also lets compounding work in your favor. A dollar that grows at a steady rate for thirty years turns into far more than most people intuitively expect, simply because growth builds on top of growth. But this only works if you are never forced to sell. Selling early, especially selling at a low point because you have to, is the single biggest way people destroy their own returns.
Why leverage is the enemy of patience
Leverage and patience cannot coexist for long. The whole point of long term investing is that you get to choose when to sell. Leverage takes that choice away from you. A lender or a broker gets to decide, based on a formula, not based on whether your reasoning about the business was right.
This is why Aschenbrenner's story is not really a story about AI stocks being a bad bet. It is a story about the danger of being right on the idea but wrong on the structure. Even Buffett, who has been enormously right about American business for seventy years, has almost never used leverage in a way that could force him to sell. He keeps huge piles of cash on hand specifically so that no one can ever make him sell a good business at a bad time. Charlie Munger put it simply: to get rich, you do not need leverage, you just need to avoid being wiped out. A small chance of a total wipeout, multiplied across a long enough career, eventually catches up with almost everyone who uses heavy leverage. That is the whole moral of the Aschenbrenner story, and it is not new. It happened to Long-Term Capital Management in 1998, it happened to countless investors in 2008, and it happened here in 2026.
Warren Buffett's big bets
There is a well known idea, often traced back to Charlie Munger, that Berkshire Hathaway's extraordinary results did not come from thousands of small, clever trades. They came from a small number of very large, very concentrated decisions, held for a very long time. Munger once said Berkshire's record was built on something like ten or twelve truly good decisions over fifty-plus years, not constant trading.
Look at the current numbers and the pattern holds up. As of early 2026, Berkshire's top five public stock holdings, Apple, American Express, Coca-Cola, Bank of America, and Chevron, made up about 67 percent of the entire public equity portfolio. Apple alone was worth more than a fifth of it. And that is before counting Berkshire's wholly owned businesses like GEICO, BNSF Railway, and See's Candies, which are not even in that public stock count but which have generated a staggering amount of the actual wealth over the decades. See's Candies, bought for $25 million, has thrown off more than $2 billion in profit over time.
The exact number, whether it is four bets or ten or twelve, is less important than the pattern. Buffett did not get rich by owning a little bit of everything. He got rich by finding a small handful of wonderful businesses, buying them without borrowed money, and then simply refusing to sell for decades. That patience, not any single stock pick, is the real skill.
Rakesh Jhunjhunwala's Titan bet
Rakesh Jhunjhunwala, often called India's Warren Buffett, built much of his legend on one stock: Titan Company, the watches and jewelry maker owned by the Tata Group. He started buying Titan around 2002 and 2003, at roughly 30 to 35 rupees a share, when almost no one else was paying attention to it.
Over the next two decades, Titan turned into one of the greatest compounding stories in Indian markets, rising well over 10,000 percent from his entry price. By the time of his death in 2022, Titan was the single largest holding in his portfolio, reportedly making up close to one-third of its total value. That is an enormous concentration for one stock to hold in a diversified portfolio, and it is a big reason his overall fortune, estimated at around 5 to 6 billion dollars near the end of his life, grew as much as it did.
It is worth being precise here, since the numbers get repeated loosely online. Titan was not literally "most" of his $5 billion portfolio in the sense of being over half. It was close to a third, which still made it by far his single biggest position and his single biggest source of gains. The rest of his fortune came from a long list of other holdings, companies like Crisil, Escorts Kubota, Star Health, and several others, plus new ventures he backed directly. But if you had to point to one decision that defined Jhunjhunwala as an investor, it would be Titan: spot a quality business early, take a real position in it, and then simply hold it through twenty years of ups and downs without flinching.
Putting it all together
Three stories, one lesson. Aschenbrenner had a good idea, wrapped it in heavy leverage, and lost most of the value in days when the market dipped and his lenders called. Buffett had good ideas too, wrapped them in patience and zero forced-selling risk, and built one of the greatest fortunes in history over seventy years. Jhunjhunwala did the same thing in India with Titan, buying early, sizing it big, and refusing to sell for two decades.
The difference between these outcomes was never really about who was smarter or who understood the future better. It was about who could afford to be patient. The investors who avoid borrowed money give themselves the right to wait out bad weeks, bad months, even bad years. The investors who lean on leverage hand that right away to someone else, a bank, a broker, a lender, who does not care about your thesis and will sell you out the moment the numbers say so.
If there is one thing worth taking from this week's news, it is this. Find good businesses. Buy them with money you actually have. Then be boring enough, and patient enough, to just hold on.
Comments