What is a stock market bubble?
There is no single agreed definition. Eugene Fama, a Nobel-winning economist who doubts that bubbles can be spotted in advance, has described a bubble as an “irrational strong price increase that implies a predictable strong decline”. Historical accounts share a pattern: a real innovation, such as radio and electrification in the 1920s or the internet in the 1990s; rising prices that pull in new buyers; borrowing to buy more; a flood of new shares; and a belief that the old valuation rules no longer apply.
The key word is predictable. A price that later falls was not necessarily a bubble, because news can change. And a price that keeps rising for years after being called a bubble was not necessarily fairly valued either. That uncertainty is what makes the subject hard, and why the rest of this lesson leans on data rather than labels.
- Stock market bubble
- A rise in stock prices far beyond what earnings can plausibly support, driven mainly by the expectation of further rises, that ends in a steep decline. It is usually recognized with certainty only afterward.
- Greater fool theory
- Buying an asset not for its earnings but because someone else is expected to pay more for it later. It works only as long as new buyers keep arriving.
The 1929 and dot-com bubbles: what happened?
In the late 1920s, US stocks rose on new industries and easy credit; many investors bought on margin, putting down only part of the price. Robert Shiller’s cyclically adjusted price-to-earnings ratio (CAPE) for the S&P index reached 32.6 in September 1929, a level it had never reached before. By June 1932, monthly average prices were about 85% lower and the CAPE stood at 5.6. Stock market crashes tells the story of the crash and the Great Depression that followed.
The dot-com bubble formed around the internet. Many young companies went public with little revenue and no profit, and buyers paid up anyway. Jay Ritter’s IPO data, in the drawing below, count 476 US IPOs in 1999 and 380 in 2000, with average first-day gains of 71.2% and 56.3%, against 14% to 22% a year in 1996–1998; 78% of the companies that went public in 1999–2000 had negative earnings, against 32% in 1990–1998. The CAPE peaked at 44.2 in December 1999, and the Nasdaq Composite, which topped in March 2000, fell about 78% by October 2002.
Notice that the number of IPOs alone was not the signal: 1996 had more IPOs than 1999. What changed was how much buyers paid on the first day, and how many of the companies had no profits at all.
- First-day return
- The percentage change from an IPO’s offer price to its closing price on the first day of trading. Very high averages across many IPOs signal heavy demand for new shares.
Was tulip mania really a bubble?
Tulip mania, in the Dutch Republic in the 1630s, is often called the first bubble: rare bulbs said to sell for the price of a house, then a collapse in 1637 that ruined the country. Historians have questioned much of that story. Anne Goldgar, who studied archives in Haarlem and Amsterdam, found that the popular account traces back to a few pieces of contemporary propaganda copied by later writers; that not everyone was trading; that many varieties never rose much in price; and that the episode did not destroy the Dutch economy or the livelihoods of most people involved. She describes it as a social and cultural crisis, about trust, more than a financial one.
Economist Peter Garber proposed fundamental, non-bubble explanations for tulipmania in 1990, and argued that economists should exhaust reasonable economic explanations before calling an episode a bubble. The lesson for investors is not that tulips prove crowds are mad, but that stories about bubbles get simpler with every retelling, which is why it pays to check the data.
- Tulip mania
- A rise and collapse in Dutch tulip bulb prices in the 1630s, often cited as the first speculative bubble. Historians now judge its economic damage to be much smaller than the legend says.
What warning signs do investors watch?
None of these signals times a top, and each has given false alarms, but together they describe how stretched a market is. The drawing at the top of the page shows the first one over 145 years:
- Valuation: the Shiller CAPE divides price by ten years of average inflation-adjusted earnings. Its average from 1881 to 2025 was about 17.7. Before 1997 it went above 30 only in August and September 1929; since mid-1997 it has spent long stretches above 30. The P/E ratio explains the basic version.
- Leverage: margin debt, the money investors borrow from brokers to buy securities, rising faster than the market. FINRA publishes it monthly from its member firms; see margin trading.
- Issuance: a wave of IPOs and new shares, especially from young, unprofitable companies, as in 1999–2000; see what is an IPO.
- Breadth and speed: a narrow group of stocks doing most of the work, and prices rising faster and faster; see market breadth.
- Stories and sentiment: crowds of new investors, talk that this time is different, and buying simply because prices are rising; see FOMO in trading and the fear and greed index.
- Cyclically adjusted P/E (CAPE)
- A stock index’s price divided by its average inflation-adjusted earnings over the previous ten years. Robert Shiller popularized it to smooth out the swings of single-year earnings.
- Margin debt
- The total that investors have borrowed from brokers against their securities. FINRA reports it monthly as debit balances in margin accounts.
Why is timing the top of a bubble so hard?
Being early looks the same as being wrong. On December 5, 1996, Federal Reserve Chairman Alan Greenspan asked in a speech how anyone would know when “irrational exuberance has unduly escalated asset values”. The CAPE was already about 27.7 that month; it went on to 44.2 three years later. Economists Robin Greenwood, Andrei Shleifer and Yang You note that Robert Shiller also warned of a bubble in December 1996, after which prices roughly doubled, and that even at the low that followed, the index stood higher than when he first warned.
Their study of US industries since 1928 adds numbers. After an industry beat the market by 100% over two years, the chance of a 40% crash within the next two years was 53%, against 20% after a 50% run-up and 80% after a 150% run-up. Yet average returns after such run-ups were not unusually low, because many industries simply kept rising. And in the 21 run-ups that did end in a crash, prices peaked on average six months after the run-up was first flagged, with about 30% more gains in between. A sharp rise raises the risk of a crash; it does not tell you when.
That is also why betting against a bubble is so risky. A short sale loses more the higher the price goes, with no ceiling, and a market that stays expensive for years can force a short seller out before any fall.
How can investors protect themselves without calling the top?
Because nobody times bubbles reliably, most protection comes from habits that help whether prices are in a bubble or not:
- Diversify across sectors and asset types, so that one crowded theme cannot sink the whole portfolio; see portfolio diversification.
- Rebalance on a schedule: trimming what has grown most and adding to what has lagged sells part of a boom without forecasting its end.
- Go easy on leverage, especially on fashionable assets: borrowed money is what turned the fall of 1929 into forced selling.
- Check the business behind the story, such as revenue, profit and cash flow, not just the chart; see margin of safety.
- For traders, use stops and position sizing, because late-stage moves can reverse violently; see risk management.