The episodes covered here stretch across four centuries, several continents, and just about every asset imaginable. What sits underneath them is remarkably narrow.
Thirty episodes get examined in this section, spanning four centuries, several continents, and nearly every category of asset anyone has ever invented, tulip bulbs and mortgage bonds, seventeenth-century merchants and present-day family offices, outright fraud and honest miscalculation. Line them up side by side and what jumps out isn't their variety. It's how few structures actually sit underneath all of them.
The first recurring structure is leverage, and it shows up in a large share of the worst outcomes on record. It brought down Long-Term Capital Management despite a roster of Nobel laureates. It brought down Lehman Brothers despite a century and a half of survival. It took Archegos down in a matter of days. It's what turned the 1929 crash from a painful loss into total ruin for anyone who had borrowed to buy in. The mechanism never changes: borrowing takes away your ability to wait, and waiting is what every sound investment thesis actually needs.
The second is concentration, and it shows up wherever a single failure was enough to be catastrophic, the Japanese investor who owned nothing but Japanese assets, the employee whose savings sat entirely in their employer's stock, the family office that staked everything on a handful of positions. Concentration isn't always a mistake. But it removes your capacity to survive being wrong, and being wrong isn't some occasional hazard in investing. It's the baseline condition.
The third is correlation, or more precisely, diversification failing at the exact moment it was supposed to work. The mortgage securities behind 2008 were built on the assumption that regional housing markets wouldn't all fall at once, and then they did. The models at LTCM assumed various positions would move independently of each other, and they didn't. In both cases the diversification had been calculated from historical data, and it evaporated under conditions the history simply hadn't included, which is exactly when it mattered most.
The fourth is narrative, and it might be the most dangerous of the four because it works on smart people through their intelligence rather than around it. The internet really was transformative. The Nifty Fifty really were excellent businesses. Japan's economy really did have remarkable strengths. In every case the story was true, and the truth of the story is precisely what let investors stop asking about price. A false story is fairly easy to resist. A true one that's already fully priced in is not.
The fifth sits underneath all the others: substituting price for value. Buyers of tulip contracts were purchasing on the expectation of reselling, not on any real assessment of worth, and buyers at the top of every bubble in this section were doing the same thing, whatever they told themselves at the time. The moment the question shifts from what is this actually worth to what will the next person pay me for it, the whole arrangement now depends on a subsequent buyer showing up. Subsequent buyers eventually stop showing up.
Sitting alongside these structural failures is a behavioral pattern that's arguably even more consistent. Investors traded too often and paid for it. They sold their winners and held onto their losers. They piled in after the good performance had already happened and got out before the recovery began. They kept far more of their own country's companies than any reasonable analysis would justify. These exact patterns turn up in brokerage data from the 1990s, and they turn up again in the meme-stock episode thirty years later, in a completely different market with completely different people.
That persistence across such different conditions is the single most important thing this section has to offer. The instruments change. The technology changes. The rules change too, usually in reaction to whatever disaster just happened. What stays constant is the person holding the instrument, and the failures catalogued here are, overwhelmingly, failures of that person rather than failures of whatever they happened to be holding.
It follows that the defenses are few and unglamorous. Don't borrow, because borrowing takes away your ability to wait. Don't concentrate to the point where one mistake would ruin you, because mistakes are guaranteed. Diversify across things that don't all depend on the same conditions. Ask what something is actually worth before asking what it costs, and refuse to own anything you can't understand. Decide in advance what you're going to do. None of it is clever, and its plainness gets mistaken for weakness far too often.
There's one last observation the whole section supports, though no single episode proves it on its own. In not one of these cases was the investor destroyed by a failure to forecast correctly. They were destroyed by a failure to survive being wrong. Many of the people who lost everything were, in fact, substantially right about whatever they were analyzing, right about the internet, right about the quality of the Nifty Fifty companies, right about the relationships LTCM had identified, right about the pandemic in early 2020. Their analysis wasn't the failure. Their structure was.
One more thing needs saying, because it would be a serious misreading to skip it: this catalogue of disaster is not an argument against investing. The same four centuries that produced every episode in this section also produced the accumulated growth that made those episodes worth studying in the first place, and an investor who reacted to all this by simply withdrawing from markets would have absorbed the warning and missed the entire point. The goal here isn't to establish that markets are dangerous, everybody already assumes that. It's to pin down where the danger actually lives. And on the evidence gathered here, it lives far less in the markets themselves than in the arrangements investors build and the behavior they allow themselves. That's genuinely good news, because it puts the problem exactly where an investor has some control over it.
This is the point the whole section is built to make. An investor's edge doesn't come from seeing further ahead than everyone else, which almost nobody can actually do. It comes from being arranged well enough to still be there when their judgment eventually gets vindicated, and from accepting that the wait for that vindication can run far longer and far more uncomfortably than they'd like. Investing better rather than faster isn't a matter of taste. It's what four hundred years of honestly examined history actually recommends.