One man turned $36 billion into $160 billion of exposure. Banks lost more than $10 billion when it unwound. His name is Bill Hwang. He ran Archegos, a family office, not a regular fund. He never needed to own the stocks. He used total return swaps, so the banks held the shares on paper and his name appeared on no public filing. Each bank saw only its own slice. None saw the full picture. On March 22, 2021, ViacomCBS announced a $3 billion stock offering. The shares fell, the margin calls came, and he could not cover them. The banks sold his positions in a rush. More than $100 billion in market value was wiped out. Bloomberg estimated Hwang lost about $20 billion in two days. The banks that moved earlier, like Morgan Stanley, avoided major losses. The ones that waited took the hit. Credit Suisse alone lost $5.5 billion. In November 2024, a federal judge sentenced Hwang to 18 years in prison. The risk was never hidden from the banks. It was only hidden from each other. Who should have seen it?

Ray Dalio went broke in 1982 and borrowed $4,000 from his father. Today he is worth about $21 billion. In 1982, Ray Dalio was sure. Mexico was heading for default, and he told clients the coming depression would be as bad as, or worse than, the thirties. He was invited to testify before Congress and to appear on Wall Street Week. He was dead wrong. The Fed loosened, stocks surged, and the U.S. economy went on to enjoy its greatest period of non-inflationary growth. Bridgewater, the firm he started in 1975, shrank to practically a one-man operation. He laid off nearly everyone, sold the second car, and borrowed $4,000 from his father to cover family bills. The lesson was not "be smarter." He stopped thinking "I'm right" and started asking "How do I know I'm right?" He hired people who disagreed with him and wrote his decision rules down so they could be tested. He also found that holding around 15 uncorrelated bets could cut risk by up to 80% without lowering returns. By Dalio's own account, Bridgewater then averaged about 11.8% a year over three decades with only minimal down years. Bloomberg puts his net worth near $21.5 billion. He calls going broke one of the best things that ever happened to him. Would you rather be right once, or be willing to be wrong?

The state has two ways to take your money. Andrew Tate faced both. In January 2023 Romanian authorities seized about $4 million of assets from Andrew and Tristan Tate: cars, properties, watches and cash. If the brothers were convicted, those assets could be forfeited to compensate alleged victims. Then the UK came in through a different door. Devon and Cornwall Police went after taxes, not the criminal case. They said the brothers earned about £21 million from online businesses between 2014 and 2022 without paying tax in any jurisdiction. In December 2024 Westminster Magistrates' Court ruled police can seize about £2.5 million from frozen business accounts. The Chief Magistrate described the structure as a "straightforward cheat of the revenue." Tate called the ruling a coordinated attack, and his lawyer said the transfers were ordinary business practice. In February 2025 a Romanian court lifted its seizure on the bank accounts, vehicles, land and some company shares. Same people, same money, two legal tools. One was reversed within two years. The other was a court ruling on tax. The brothers deny all charges, and the Romanian criminal case has no trial date. When your income is fully online, who decides which country gets to tax it?

Elon Musk is renaming SpaceX’s AI division from SpaceXAI to SpaceXSI. The move follows Trump’s push to call artificial intelligence “super intelligence.” Musk confirmed on X: “Yes, we will make that change” and added “No more AI. SI, it’s better.”

Most billionaires fight over inheritance. Durov designed his to prevent the fight. In June 2025 Pavel Durov told Le Point he has 106 children in 12 countries. Six are biological, with three partners. The rest came from sperm donation he began about 15 years ago. His answer to the obvious question was simple. They all have the same rights. On a $13.9 billion fortune, that is roughly $130 million each. The catch: they cannot touch it for 30 years. His reasoning was about character, not tax. He wants them to build themselves up alone, "not to be dependent on a bank account." And he does not want them tearing each other apart after he is gone. Notice the design. Equal shares remove the reason to compete. The 30 year lock removes the reason to wait. Then the number moved. In March 2026 Forbes cut his net worth to $6.6 billion. Same rule, 106 heirs, now about $62 million each. The fortune is not fixed. Only the rule is. Would you take $130 million in 2055, or build your own first?

Two Nobel Prize winners ran a fund with 25x leverage. It lost $4.6 billion in under four months. Long-Term Capital Management started trading in February 1994 with about $1 billion. Myron Scholes and Robert Merton were on the team, and in 1997 they shared the Nobel Prize in Economics. After fees the fund made roughly 21%, 43%, and 41% in its first three years. Then it did what winners do. It added leverage. By early 1998, $4.7 billion of equity was supporting $129 billion of assets, a debt-to-equity ratio above 25 to 1. Off balance sheet derivatives had a notional value near $1.25 trillion. Leverage turns a small mistake into a fatal one. The models said the spreads would converge. In 1998 they did not. On September 23, 1998, the New York Fed gathered 14 banks to put in $3.625 billion. Not because one fund mattered, but because its trades touched everyone. The fund was liquidated in early 2000. Smart is not the same as safe. LTCM did not fail because its math was wrong most of the time. It failed because it was wrong when it was 25 times levered. Which would you rather have, a great model or a small position?

Before you get excited about the new $5,000 check, here's what happened to the last two. February 2025: a $5,000 "DOGE dividend" floated, originally investor James Fishback's idea, Trump said he loved it. Never paid. November 2025: "at least $2,000" tariff dividend, White House called it a commitment. Never paid. September 2026: now it's $5,000 again, branded the "Trump dividend," conditional on Republicans holding Congress on November 3. Meanwhile the Supreme Court struck down the underlying IEEPA tariffs in February, importers are owed roughly $166B in refunds, tariff revenue is under $200B a year against a $2T deficit and $1T in annual interest payments. The Fed just hiked rates in September for the first time since 2023, and the 10-year yield is sitting above 5%. In 2020 the government could print its way into stimulus checks while the Fed's balance sheet nearly doubled. This time the Fed is tightening, not expanding, and the tariff pot it would pay from is already underwater. $5,000 landing in your account next year, funded by money the government doesn't currently have. Third time the charm, or third time nothing shows up?

What happens to a habit when the study behind it gets retracted? For years I've cut every big project into small deadlines I set for myself. They live on a sticky note at the bottom of my monitor, one line per Friday, like a chore chart on a fridge. Part of the reason was a 2002 paper by Dan Ariely and Klaus Wertenbroch. Self-imposed deadlines help, evenly spaced ones help more. I had assumed that part of my week was settled science. This September the journal retracted it, at Wertenbroch's own request. The data sleuths at Data Colada found that in one experiment, 18 of the 20 people in one group had a duplicate elsewhere in the data, down to identical proofreading scores. Ariely's own advice about the finding is now to believe it less. Turns out the habit didn't break. That's the annoying part. I looked at the sticky note after reading the notice and the Friday lines were still there, still working on me. The study was never holding the habit up. It was the receipt I kept in the drawer after I'd already bought the thing. What still stands is smaller. People really do choose to bind themselves early, and that part has held up. Whether evenly spaced deadlines make the work better is another matter, and a new replication in the same journal found no evidence that they do. Ulysses figured out the binding part long before any lab did. He had his crew tie him to the mast so he could hear the sirens without being able to sail toward them, and nobody ran a controlled trial on the rope. The sticky note is still on the monitor. This Friday's line says finish the draft, and I haven't decided yet whether I believe it.

In October 2008, Porsche revealed it quietly controlled 74% of Volkswagen, and by the next afternoon VW was the most valuable company on earth. The disclosure came on October 26. Porsche said it directly owned 42.6% of VW's ordinary shares and held cash-settled options on another 31.5%, adding up to 74.1% of the company. Lower Saxony's government owned another 20%. That left less than 6% of VW shares actually available to trade. Hedge funds had shorted roughly 12% of the company, betting the stock would fall. There was nowhere near enough stock left for them to buy back and close those positions. VW's share price went from about 210 euros to over 1,000 euros in less than two trading days. On October 28, the stock peaked at a market value of 296 billion euros, pushing VW ahead of ExxonMobil as the most valuable public company on the planet, for a few hours, before the price came back down. Estimated losses for the short sellers: $30 billion. Porsche never had to buy a single additional share to make it happen. It just had options already in place and let the math do the rest. Germany's regulator opened an investigation into possible market manipulation. Porsche denied it, then sold off about 5% of its stake the next day to calm things down. Is a trade still a short if there's nothing left to buy back?

Two billionaires went on live television in January 2013 and tore into each other over a nutrition company most people have never shorted or owned. Bill Ackman had put $1 billion into betting Herbalife would collapse, calling it a pyramid scheme in a 342-slide presentation. Carl Icahn took the exact opposite side, buying in heavy, and when CNBC caught him on the phone he opened with "He's like the crybaby in the school yard." Ackman shot back that Icahn "takes advantage of little people." Icahn closed it with "I wouldn't invest with you if you were the last man on Earth." None of this was really about Herbalife. The two men had been feuding since a soured 2003 real estate deal called Hallwood Realty, and this was just the public eruption of a decade-old grudge. Ackman held his short for five years. He finally closed it out in February 2018 with Herbalife trading near $92, more than double his entry point. Icahn walked away with roughly $1 billion in profit and didn't fully exit his stake until 2021. The guy with the 342 slides and the regulatory theory lost. The guy settling an old score won.

Knight Capital trading glitch (2012): On August 1, 2012, a single line of code — dead since 2003 — woke up for 45 minutes and cost a Wall Street firm $440 million. Knight Capital's order system had a feature called "Power Peg," retired nine years earlier. Nobody deleted the old code — it just sat there, unused. In July 2012, engineers reused that same flag for a new program. When they rolled out the update, one of their servers failed to receive it. Silently. No alert, no error. So on the morning of August 1st, that one server started reading new orders as if they were Power Peg instructions from 2003. It began buying and reselling the same stocks over and over, accumulating positions nobody had asked for, at a rate no human could watch in real time. Between 9:30 and 10:15 a.m., the system built up 397 million shares across more than a hundred stocks — about $7.65 billion in unintended positions. By the time anyone pulled the plug, the firm was down $440 million. More than the company's entire market value a day earlier. Four days later, Knight took a $400 million rescue just to survive the weekend. It never really recovered — merged into KCG a year later, then sold off piece by piece until Virtu Financial owned what was left.

THE ROADSTER IS FINALLY HERE. 🚀 Tesla is finally showing what fans have been waiting for years. Insane design. Electric power. And technology that could make it the fastest Tesla ever. Is it finally happening?

George Soros / Black Wednesday (1992): On September 16, 1992, the Bank of England raised interest rates twice in a single day — and still lost. The UK had pegged the pound to the German mark under the Exchange Rate Mechanism, at a rate most traders thought was too high to hold. George Soros's Quantum Fund built a short position against the pound worth roughly $10 billion, betting the peg would break before the Bank's reserves would. By 11 a.m. that day, the government hiked rates from 10% to 12%, trying to make holding pounds more attractive than selling them. It didn't work. A few hours later, rates went to 15% — a second emergency hike in the same morning. The Bank was also buying roughly £2 billion of sterling an hour, burning through reserves in real time. None of it held. At 7:40 p.m., Chancellor Norman Lamont stepped outside the Treasury and announced Britain was leaving the ERM. The defense had cost the UK government an estimated £3.4 billion. Soros personally walked away with over £1 billion — made in a single day, betting against a central bank with the full reserves of a G7 economy behind it.

Lehman Brothers collapse (2008): Lehman Brothers didn't have to collapse. There was a buyer at the table, the deal was basically done, and it died over one rule a British regulator wouldn't waive for 48 hours. That weekend, Treasury Secretary Hank Paulson refused to put public money behind another rescue — Bear Stearns had already used up the political room for that six months earlier. So the Fed pulled the heads of every major bank into a room and pushed them toward buying Lehman themselves. Barclays was the frontrunner, and by Saturday the terms looked close to final. Then the UK's Financial Services Authority stepped in. British law required shareholder approval for an acquisition that size, and the regulator wouldn't grant an emergency exemption just because Wall Street was unraveling three thousand miles away. No waiver, no deal. By Sunday night, Lehman employees were already clearing out their desks. On September 15, 2008, Lehman filed for Chapter 11 with over $600 billion in assets — still the largest bankruptcy in US history, before or since. Two days later, Barclays came back anyway — not for the whole firm, just the parts worth having. $1.75 billion for the investment bank, the Seventh Avenue headquarters, and two data centers, stripped of all the toxic assets that sank the rest. When the ink dried, a Barclays executive played "God Save the Queen" over the office intercom.

In 1999, a fraud investigator named Harry Markopolos handed the SEC a memo proving one hedge fund's returns were mathematically impossible. He warned them again in 2000, 2001, 2005, and 2007. Five warnings, nine years, two different SEC offices. Nobody opened the file that mattered. The fund manager was Bernie Madoff. On paper, his firm never had a losing month — a steady 10–12% a year, in bull markets and bear markets alike, for two decades. No legitimate strategy produces that. Markopolos worked it out in about four hours. What actually ended it wasn't the SEC. It was 2008. The crash triggered roughly $7 billion in redemption requests Madoff couldn't pay, because there was no fund behind the returns — just new client money paying off old clients, the oldest trick there is, run at a scale nobody thought possible. On December 10 he told his sons the business was "one big lie." They called a lawyer that night; he was arrested the next morning. The number everyone remembers is $65 billion — the fictional balance sitting on client statements. The number that actually mattered was closer to $17–18 billion, the real cash investors put in and lost. Both are true. They're not the same fraud. He got 150 years. Markopolos got nothing — no bonus, no promotion, just five ignored letters sitting in an SEC file for nine years.

This isn’t a chatbot anymore. It’s an employee that works while you sleep. OpenAI just launched dots on GPT-6 Astra — an agent with its own computer, 4,000+ apps, and the initiative of a chief of staff. It can book a table or run a serious project in the background, 24/7. Control is supposedly still yours. Would you give it access to your email today?

In January 2008, a bank spent three days secretly unwinding a $50 billion position that one of its own junior traders had built without anyone noticing. The unwinding itself helped crash the very market it was trying to escape. Jérôme Kerviel joined Société Générale's compliance department in 2000, then moved to a junior trading desk in 2005. Starting in late 2006, he began building unauthorized positions in European stock index futures, hiding them behind fake offsetting trades he closed out every two or three days — just before the bank's internal alarms would trigger. By the time Société Générale's own risk systems caught him, on January 19, 2008, his hidden position had grown to nearly €50 billion — more than the entire market value of the bank that employed him. Along the way he'd apparently generated over a billion euros in paper profits; his bonus that year was expected to be €300,000. The bank didn't disclose what it had found. Instead, over the next three trading days, it quietly sold off the entire position into markets that were already falling. By the time the unwind was done, the loss stood at €4.9 billion — and some traders have argued the forced selling itself deepened the global rout that hit markets on January 21, days before the Federal Reserve made a surprise rate cut. Kerviel was convicted in 2010 and ordered to personally repay the full €4.9 billion — a sum neither he nor almost anyone on Earth could produce in a lifetime. In 2014, France's highest court threw that repayment order out entirely, ruling that Société Générale's own negligence had contributed to the scale of the losses. The bank that built its case on one man's fraud couldn't collect a cent from him once a court actually looked at what it had missed.

In May 1986, a Wall Street trader stood in front of UC Berkeley's graduating business students and told them greed was healthy. Six months later, he was the reason federal regulators announced the largest insider trading settlement in Wall Street history. Ivan Boesky built his fortune on risk arbitrage — betting on which companies would get taken over before the deals became public. He started in 1975 with $700,000 borrowed from his wife's family. By 1986 he was worth more than $200 million, including $136 million from selling the Beverly Hills Hotel alone. At that Berkeley commencement, Boesky told the crowd: "I think greed is healthy." The line stuck. The next year, Oliver Stone's Wall Street gave Gordon Gekko a version of the same idea, and audiences are still quoting it decades later. Boesky wasn't a screenwriter's invention making a point about capitalism — he was describing, almost word for word, how he actually ran his business. On November 14, 1986, the SEC announced it had settled insider trading charges against him. He'd been paying investment bankers for advance word on mergers before the public ever saw them. Boesky pleaded guilty to conspiracy, paid a then-record $100 million penalty, and agreed to secretly record his own phone calls with contacts — including junk bond financier Michael Milken — for federal investigators. Boesky served 20 months of a three-and-a-half-year sentence. The line he gave away for free at a graduation ceremony outlived the fortune, the firm, and very nearly the freedom that built it.

A bank that survived Napoleon, financed the purchase that doubled the size of the United States, and banked British royalty for two centuries — was wiped out in under three years by one 28-year-old trader working alone in a back office in Singapore. Barings Bank, founded in 1762, was Britain's oldest merchant bank. It underwrote the 1803 Louisiana Purchase and financed the U.S. government through the War of 1812. By the early 1800s it was informally counted among the great powers of Europe. Generations later, the Baring family line produced a great-grandmother of Princess Diana. In 1992, Barings sent Nick Leeson — who'd started his career as a clerk at Coutts, the Royal Family's own private bank — to open a derivatives desk in Singapore. There, Barings let him run both the trading floor and the settlements office meant to check his own trades. No one was positioned to catch him. Leeson hid a growing pile of unauthorized bets on Japanese markets inside an obscure internal error account, number 88888. When the January 1995 Kobe earthquake sent those markets into freefall, the hidden position collapsed with them. By the time London found out, the account had swallowed £827 million — more than the bank's entire capital. Leeson fled Singapore leaving a two-word note: "I'm sorry." He was arrested in Frankfurt months later. Barings, 233 years old, was insolvent within days. ING, a Dutch bank, bought the whole institution — its name, its royal clients, its 233 years — for £1. The world's oldest merchant bank didn't die because someone made a bad bet. It died because nobody was watching the person making it.