What the Crises Have in Common

Why this matters

No two crises are alike, and the next one won't copy any of these. But the twelve episodes in this module, from The Panic of 1907 to The 2022 Stock and Bond Decline, share a small set of mechanisms. Knowing them won't tell you when the next crisis comes. It can tell you what makes a portfolio, a firm, or a financial system fragile, and how to avoid being forced to sell at the worst moment.

1. Borrowed money

Leverage turns a price fall into forced selling. In the 1920s, buyers of stocks typically put down 10% and borrowed the rest, so a 10% fall wiped out their stake.¹ Long-Term Capital Management held over $125 billion of assets on less than $5 billion of capital,² and Orange County's pools borrowed about $12.6 billion to hold $19.8 billion of securities.³ Before 2008, household mortgage debt rose from 61% to 97% of GDP.⁴

Leverage doesn't cause crises by itself. It removes the option of waiting. An unleveraged investor can sit through a fall; a leveraged one may be sold out at the bottom, and those sales push the price down for everyone else.

2. Runs on short-term money

The institutions that fail first are usually those that fund long-term assets with money that can leave overnight, and that sit outside the safety net. In 1907 it was New York's trust companies, which held about 5% of deposits in cash, against 25% at national banks, and weren't members of the bankers' clearing house. In 2008 it was investment banks and the "shadow banks" that lent to them overnight.⁵ When a prominent money market fund couldn't keep repaying shares at $1 in September 2008, investors pulled money from similar funds.⁶ A run is about cash today: in 1907, Knickerbocker Trust closed before anyone could tell whether it was solvent.⁵

3. Selling that feeds on itself

Many crises include a rule that makes sense for one investor and becomes dangerous when everyone follows it. Portfolio insurance sold futures as prices fell, and its selling triggered further rounds of selling in 1987.⁷ In 2010 an algorithm sold faster as volume rose, while high-frequency traders passed contracts back and forth.⁸ Margin calls in 1929 and LTCM's feared fire sale in 1998 worked the same way. The common lesson: when many participants hold the same position or follow the same rule, the exit is narrower than it looks.

4. "It can't fall"

Most bubbles rested on a belief that recent history would continue. The Bank of Japan described a "myth of ever-rising land prices" in 1990.⁹ Before 2008, large nationwide falls in US home prices had been rare, and many gauges of mortgage quality were built on older, safer loans.⁴¹⁰ LTCM's losses greatly exceeded what risk models estimated in calmer periods suggested were probable.² Orange County's strategy assumed rates would stay low for at least three more years.³ A long run of calm is evidence about the past, not a guarantee about the future.

The rest of this lesson is for members

Still to come in this lesson: 5. The central bank, 6. What it meant for investors, the key takeaways, and 2 quiz questions with explanations.

Membership opens every lesson in the curriculum, the capstone projects, the interactive tools, and the members-only channels in the Hyperion Entities Discord. The first lesson of every module stays free.

Become a member Sign in