Famous first bubbles

Stories about a number of famous historical "bubbles" continue to be repeated in popular discussions of so-called manias and presumed irrational investor behavior. However, it is important to understand that the mere repetition of folklore does not prove the existence of actual bubbles, nor can it prove that the well-known old interpretations of historical events are correct. It is also important to remember that many authors have a strong bias towards one of two basic views: 1) markets (in particular financial markets) are inherently bad and people are by nature emotional and irrational (psychological), or 2) markets with self-interested participants are the best way we know to organize economic trading and self-interested market participants, on average, need to behave in a rational manner. These two fundamental biases are bound to persist, because, despite our best attempts, economic and statistical research has shown to be unable to provide definitive answers to our questions and results are almost always open to alternative interpretations depending on particular assumptions.

The basic economic fundamentals of all (durable good) asset prices are expectations about the discounted value of future earnings from the investment. It serves no purpose to designate an observed pattern of price increase and collapse as a bubble, when we know only with the benefit of hindsight that optimistic expectations did not materialize (i.e. ex post). Assets can be rationally priced even when investors make rational but imperfect forecasts and when, unfortunately and sometimes disastrously, optimistic expectations do not materialize. Second, we need to remember that even when forecasts are perfect and markets are efficient, asset prices can go up and down in predictable patterns without generating opportunities for excess profits. The longstanding idea, unfortunately still perpetuated by many economists, that, in efficient and rational markets prices should be random walks (i.e. unpredictable, martingale processes) is simply demonstrably wrong.

A "bubble" in asset prices is only a bubble in the true meaning of the concept when asset prices are clearly seen to deviate from fundamental values that could rationally have been expected at the time of trading (i.e. ex ante). A bubble is "a market-determined price at odds with any reasonable economic explanation". Obviously, 'beyond reason' meaning in this context something different than the very legitimate difference of opinion that can occur between two or more groups of people.

In this context of the real definition of a financial market bubble, lets re-examine some of these so-called famous first bubbles, from a perspective that rational investors might have used.

Note: There is, of course, no denying the very obvious fact that through time and across countries, some individuals (sometimes even groups of individuals) behave in very silly, stupid and outright irresponsible ways. There seems to be a never ending supply of individuals who, despite the warnings of history, continue to bet on unlikely investment schemes that turn out to be fantasies, ponzi schemes, chain letters, pyramid schemes, etc. Sometimes, in special cases, we can regard these bets as unique attempts to escape from an already hopeless economic position (referred to as 'rational gambling' on outlier payoffs). However, most of the times we cannot. At the same time, note that economics as a science is generally about understanding and explaining the average behavior of average economic agents (firms, consumers, policymakers). Economics cannot, does not and needs not attempt to explain the behavior of every individual at every moment in time. In short, the behavior of markets cannot and should not be equated with the behavior of some individuals.

So-called famous bubble episodes discussed on my site are (see the menu links, left):

• Tulipmania 1637
• South Sea Bubble 1720
• Mississippi Bubble 1720
• Dutch Mirror of Folly, Windhandel 1720
• Stock Market Crash of 1929 and the 1920s Technology bubble
• Stock Market Crash of 1987 and the the 1980s Computer Technology bubble
• Stock Market Crash of 2000 and the 1990s Internet bubble

Special note: In the literature there exist several empirical approaches to testing for asset price bubbles. One type of empirical test for bubbles has gained a large following in the academic literature. However, its empirical application is based on a methodological fallacy.

• The Variance-Bound Fallacy

Most other tests suffer from similar weakness in respect to application and interpretation. Tests for asset price bubbles such as random walk tests are simply irrelevant because wrong.

Empirical models and tests of fundamental value on the other hand frequently suffer from excessive simplification and lack of adequate data.

John Cochrane's review of Peter Garber's (2000) Famous First Bubbles (Word file)

Charles Kindleberger's review of Peter Garber's (2000) Famous First Bubbles (Word file)

Tim Harford (2006), Buy! Buy! Buy! Sell! Sell! Sell!: The rational explanation for stock market frenzies and crashes, Slate: The Undercover Economist, http://www.slate.com/id/2143025/

Lee (1998), Market crashes and informational avalanches, Review of Economic Studies, vol.65 ( ) : 741-59.

Barlevy (2007), Economic theory and asset price bubbles, FRB Chicago Economic Perspectives, vol. 31 (3): 44-59.

O'Hara (2008), Bubbles: Some perspectives (and loose talk) from history, &nbsp, Review of Financial Studies, vol.21 (1): 11-17.