Start anywhere. The mistakes repeat, the numbers do not.
A congressional report repriced a broker. An Ethereum upgrade shipped to the test network and a 20x long shipped to zero. The bond auction everybody was braced for went fine. And SpaceX told its shareholders the AI buildout will be paid for in bonds.
The index funds had finished buying on Monday. A price target was reached in the first minute and never seen again. A plague scare was assessed and downgraded. And a stock went to a record high on information that had been public since June.
An election surprised, a vaccine trial hit every endpoint, a logistics company bought a rival at a 29% premium, and an investor put $37.5 billion on a division nobody can buy yet. Four traders were standing on the far side of all of it.
The weakest payrolls print of the cycle sent the Nasdaq 100 to a record. Four traders lost $138,097 to a press release, a flood and no news.
The August inflation beat was the size of a measurement change the government had announced in advance. Bonds worked it out by lunchtime. Boeing won a $20bn fighter contract and closed down. Northrop lost it and closed down more.
The housing regulator let a rival credit score into Fannie and Freddie's pricing grid and Fair Isaac fell 26.52%. A gene therapy missed at 48 months and fell 37.38%. Crude settled down 3.48% because a pipeline restarted.
MongoDB lost 18.46% because its chief executive took a job at Meta. Boeing lost 6.91% on a software bug. December gold lost $152.80 an ounce, and a miner lost twice that in percentage terms.
Twilio fell 7.96% and is still above its average price target. A pool retailer heading into Chapter 11 carries a $2.38 target last updated in May. Four traders lost $122,388 in the market's best week in three.
The ten-year Treasury yield closed at 5.106%, the highest since 2007. McDonald's promised to spend $8.5bn through 2036 and fell 4.81%, Paychex fell 8.77%, and the largest long-bond fund printed the lowest close in its history.
The S&P 500 finished flat while financials closed at a July low. A gym chain lost 9.46% because an app can cancel memberships, a biotech lost 11.59% on a trial that worked, and Medicare published a price list.
A possible meeting took $4.52 a barrel out of crude, an email-sorting app put 12.14% on Intel, Novo's plan for 2035 cost 7.96% in an afternoon, and HP declined to forecast next year at all.
A biotech filed the application for its lead drug and fell 30.69%, a generator company was repriced as AI infrastructure on a $2.4 billion Amazon deal, and a reactor developer handed back a rally that a bill in the House had given it.
The Fed raised rates 25 basis points on a unanimous vote, exactly as priced. A trucker lost 13.30% to diesel at a record $6.31, a curve trade lost money because the hike worked, and a short lost to a launch date. None of the three was really a bet on the Fed.
A lender told its shareholders it did not need to become a bank and lost 23.43% anyway, a company borrowed a billion dollars at a 0% coupon and fell 9.81%, and an arcade chain called a 1.6% sales decline an improvement. The 10-year closed above 5% for the first time since 2007.
An AI safety essay took 5.25% off Micron and put 13.85% on CrowdStrike on the same afternoon, and a drone strike on a Saudi pipeline moved Brent more than it moved West Texas. Three people owned the wrong half of each.
A man who locked in 5.28% on the thirty-year lost three months of coupon in one afternoon, a bitcoin holder found out that the ETF which made it institutional is also the pipe the money leaves through, and a man who called the inflation print to the decimal watched the S&P 500 rally 0.86% on it.
The Bancorp fell 22.33% because its own customer bought a bank charter, Tyra fell 17.66% because one patient in fourteen went the wrong way, and a man who held Booking for a three-year appeal lost it in one afternoon in Luxembourg.
Novartis fell 14.14% on its third trial setback in a week, the company that licensed it one of the failed drugs fell with it, and a man short Intel into a $20 billion share sale watched it rise 9%.
A crude short that read the word deal, gold bought as war insurance that fell because the war is inflationary, and a yen carry trade re-entered on the theory that the shock was priced.
Another guidance cut and an eight year low, an event whose news was the absence of news, a breakout that stopped at the exact level everyone was watching, and 162,000 jobs against a consensus of 55,000.
A carry trade that met a hawkish Bank of Japan, a company that beat every estimate and fell 13.54%, a $3.65 billion backlog attached to $33 million of revenue, and a stock that beat by $60 million and missed by $50 million.
A gold hedge doubled at the worst possible moment, a third of an account on one lupus readout, 215 times earnings meeting a rising yield, and a short in a dying company that got bought for cash.
A utility repriced by a legislature that went home, a crude short caught by a weekend strike, an airline that pays for the same barrel, and a moat dissolved by one Saturday post.
A short that was right about the arithmetic and wrong about the month, a currency trade that met a brand new Fed chair, and a takeover that stopped being one.
A 35.67% session at a retailer whose comparable sales went nowhere, $17 billion of bad news that made a stock go up, and a short that survived the press release but not the earnings call.
A record one day collapse at a retailer whose own stores were fine, a 3% hit from a company he did not own, and a stock repriced by one analyst's spreadsheet.
A lawsuit that beat an earnings model, a tariff four months away that got priced today, and $3 billion of shorts on the wrong side of one rally.
A dilutive overnight share sale, a 30x bitcoin long bought at the top, and a trade replayed one war too late.
A 177 percent single day squeeze, a beat that got sold anyway, and a bond trade undone by one Treasury announcement.
A chip stock massacre with a retail beat hiding inside it. Four people picked the wrong side of Tuesday.
Four real market moves this week. Four people leaned the wrong way on every one of them.
Four people spent this week paying for something they assumed would keep behaving. Software, the consumer, the range, the thesis.
A bridge that passed multiple audits accepted deposits that never existed. That was only the second worst idea this week.
A free call, a borrowed short, a tiny index move and a barrier erased $345,870 from four composite accounts.
A takeover, a dilution plan, an oil spike and one quiet earnings move erased $212,040 from four composite accounts.
Four traders discovered that being early, leveraged and certain is just a faster way to be wrong.
Three of today's four losers were right about the analysis. All four were wrong about leverage.
The Financial Darwin Awards is your portfolio's support group. We collect the internet's most spectacular money losses so you can make different mistakes.
Five minutes of reading. A lifetime of bad decisions to study. Your portfolio will thank you, eventually.
