Trading and Exchanges Chapter 28: Why Bubbles Form and Markets Crash (Part 1)
This is the most dramatic chapter in the entire book. Bubbles inflate, crashes wipe out fortunes, and panic replaces logic. If you ever watched a stock chart go vertical and wondered “how does this end?”, Harris answers that question here. Spoiler: badly for whoever is holding the bag last.
The Positive Feedback Loop
Harris starts with what he calls the price accelerator. When prices rise, optimists who bought early get richer and buy more. Their buying pushes prices even higher. Meanwhile, pessimists who shorted are losing money. Brokers issue margin calls, forcing them to buy back their shorts. That buying also pushes prices up.
Both sides end up pushing prices the same direction. Winners buy because they feel smart. Losers buy because they have no choice. This is a positive feedback loop, and it is the engine behind every bubble.
How Bubbles Form and Pop
Bubbles start with real excitement about something. Radio in the 1920s, the internet in the 1990s. The technology is real. The problem is traders cannot tell when prices already reflect the good news. They keep buying into a price that has already priced in the future.
Momentum traders pile on. Your neighbor made 40% last month, now you want in. Order anticipators front-run the incoming buyers. Prices accelerate.
Value traders who see the overpricing often cannot stop it. They might not have capital for large short positions. They might not be confident enough to bet against the crowd. Even if you are right about a bubble, being early on a short means losing money until proven right. Your broker might force you out before that happens.
Eventually sellers get aggressive. Late buyers panic. Margin calls force selling. Stop losses trigger. The same feedback loop that pushed prices up now pulls them down at double speed.
Harris emphasizes: crashes often start with bad news, but the news is almost never big enough to justify the drop. The bad news is just the match. The fuel was the bubble itself.
The 1929 Crash
October 28-29, 1929. The Dow dropped 13% then 12% on consecutive days. It bounced 12% the next day, but that was a trap. By July 1932, the Dow sat at 11% of its peak. Eighty-nine percent of value gone. Took 25 years to recover.
Classic setup. From 1924 to 1929, the Dow rose 300%. Speculators borrowed heavily. Everyone was excited about radio companies, the tech stocks of that era. RCA did not reach its 1929 high for 34 years. When prices slid, margin calls cascaded. Forced selling triggered more selling. Panic fed on panic.
Side note: most economists say the crash did not cause the Great Depression. They blame tight monetary policy and banking failures in 1930-31. But the crash got Congress to give the Fed authority over stock margins. Minimum margin was set at 45%.
The 1987 Crash: Portfolio Insurance Gone Wrong
October 19, 1987. The Dow drops 23% in one day. Still the record. The cause sounds absurd: portfolio insurance.
Portfolio insurance was a dynamic trading strategy based on Black-Scholes. Instead of buying actual put options, fund managers replicated puts by automatically selling stocks when prices fell. Sounds clever until everyone does it simultaneously. By 1987, about $100 billion was covered. A 1% market drop required selling 10 million shares against daily volume of 160 million. The math was terrifying.
When the market dropped 5% on Friday, everyone knew Monday would be bad. It was worse. Selling overwhelmed the NYSE. Dot matrix printers on the floor could not print orders fast enough. Some broke. Orders sat in queues for over an hour. Traders who could not confirm execution sent duplicate sell orders by phone. Nasdaq dealers took their phones off the hook.
The S&P 500 futures dropped faster than cash stocks, opening a 10%+ gap. Arbitrageurs who normally keep them connected stopped trading because they could not get reliable executions. Markets became disconnected.
The twist: unlike 1929, it recovered fast. New all-time high within two years. For the full year 1987, the Dow was up 2%. Panicked sellers on October 19 made a terrible decision. It was a technical crash, not a fundamental one.
Quick Stories: 1989 Mini-Crash and Palladium
The 1989 mini-crash is bizarre. Market dropped 7% one Friday afternoon after a failed UAL buyout. Many traders had already left early because the weather was nice. Market was thin, bad news hit at the worst moment. There is even an unconfirmed theory that some traders deliberately bluffed the market down, exploiting the low liquidity near the second anniversary of 1987.
The palladium cold fusion story is different. When Fleischmann and Pons announced cold fusion in 1989, palladium futures spiked 24%. Labs failed to replicate, price crashed back. Harris argues this was actually rational. If cold fusion worked, palladium would have been enormously valuable. Bad bet, not a crazy one.
The Pattern
It is always the same. Excitement builds. Leverage amplifies. Feedback loops push prices past fundamentals. Then a trigger reverses everything, and the same loops that inflated the bubble now accelerate the crash. Margin calls, stop losses, panic selling, system overloads. A crash is just a bubble running in reverse at double speed.
In Part 2, Harris covers what regulators try to do about this: circuit breakers, trading halts, and whether any of it actually works.
Previous: Chapter 27: Floor vs Automated Trading
Next: Chapter 28: Circuit Breakers and Market Safeguards (Part 2)