The Art of Contrarian Trading Part 2

INFORMATION CASCADES INTO THE WHIRLPOOL OF SPECULATION.

Let's now examine the process by which a collective wisdom equilibrium in the stock market can dissolve in a whirlpool of speculation. Surprisingly, economists have some important things to say about these departures from collective wisdom.

Suppose we look closely at the behavior of investment crowds. They share one important characteristic: The members of every investment crowd are certain that the crowd's size is evidence of the correctness of the crowd's beliefs. So many people can't be wrong! An investor outside the crowd is impressed not just by the crowd's investment success but by the unanimity of its beliefs and expectations. Experience reinforced by an instinctive belief in collective wisdom then causes him to put aside his skepticism. He believes the crowd's members must collectively know more than he does. In this way the crowd gains one more member.

But it is the nature of an investment crowd that its shared beliefs are not simply the average the independent beliefs of the crowd's members. Remember that an investment crowd grows because outsiders prefer to replace their own beliefs with what they think are the more accurate and wise beliefs of the crowd. In this way the crowd's shared belief is like a cosmic black hole. It is a powerful conviction, usually supported by well-known facts, which is an irresistible attraction for individuals and which overwhelms their personal knowledge and experience.

Surprisingly, this behavior need not be irrational on its face. It has been studied by economists. It goes under the name of an information cascade information cascade . The original source for this idea is a paper by Sus.h.i.+l Bikhchandani, David Hirshleifer, and Ivo Welch (BHI) that was published in the . The original source for this idea is a paper by Sus.h.i.+l Bikhchandani, David Hirshleifer, and Ivo Welch (BHI) that was published in the Journal of Political Economy Journal of Political Economy in 1992 (pp. 992-1026) under the t.i.tle "A Theory of Fads, Fas.h.i.+on, Custom, and Cultural Change as Information Cascades." in 1992 (pp. 992-1026) under the t.i.tle "A Theory of Fads, Fas.h.i.+on, Custom, and Cultural Change as Information Cascades."

Simply put, an information cascade is a situation in which an individual imitates the behavior of others without regard to his own information. He does so because he believes that his own information is inferior to the information of those he chooses to imitate. The idea of successive imitation, that a cascade is a sequence sequence of choices made over time by different people, is important here. And it is the sequenced cascade of imitation that I believe chronicles the life of any investment crowd. of choices made over time by different people, is important here. And it is the sequenced cascade of imitation that I believe chronicles the life of any investment crowd.

Why should an individual choose to imitate the actions of others? Sociologists have studied this phenomenon carefully. Among the explanations they offer are the preference for conformity (which has high survival value for an individual who wishes to maintain strong social bonds), as well as the fact that a strong group can enforce sanctions on deviants. But as BHI point out, the sociological explanations aren't sufficient to explain why ma.s.s behavior is so often fragile is the sense that small shocks can result in big changes.

Information cascades are fragile precisely because the collective information contained in the cascade can be seen as a pyramid standing upside down on its point or vertex. A very little information held by a few individuals right at the start of the cascade induces a great number of individuals to abandon their own collectively more substantial information and instead imitate the early innovators. Any small piece of information that subsequently is seen by some member of the cascade to contradict the information upon which the cascade is built can cause the entire structure to tip over and collapse.

Here in the theory of information cascades we again encounter the theme of fragility in market and group behavior. Changes in collective behavior and ma.s.s psychology can occur rapidly and for no obvious reason. The best way to think of this phenomenon is to compare it to the spread of an epidemic. A wonderful discussion of epidemic-related aspects of collective behavior can be found in the book The Tipping Point The Tipping Point by Malcom Gladwell (Little, Brown, 2000). His book is subt.i.tled by Malcom Gladwell (Little, Brown, 2000). His book is subt.i.tled How Little Things Can Make a Big Difference How Little Things Can Make a Big Difference, and it has a lot to say about the reasons people imitate each other, each ill.u.s.trated by illuminating examples.

Let's see how the theory of information cascades sheds light on the behavior of investment crowds. In the theory of cascades, a person imitates the actions of those who have preceded him in the cascade because he believes that they know something he does not know. Moreover, this belief in the superiority of other people's information can be an objectively correct one, a belief founded on the best statistical use of one's own information.

An investor who is not yet part of an investment crowd generally knows a number of things. He knows that the crowd is focused on an a.s.set whose price has changed a lot in a relatively short period of time. Indeed, this price change may well be what has attracted our hypothetical investor's attention in the first place. Our investor also knows the beliefs and forecasts shared by the crowd, which he has learned from its current members. To be persuasive, these must be grounded in facts that are public knowledge. There has already been a big price change in the a.s.set, and this lends more credence to the crowd's beliefs. Yet another element of persuasion may be the financial gains made by friends and acquaintances of our investor who are already members of the crowd.

It is not surprising that in such a situation an investor may rationally choose to make the crowd's beliefs his own and join the information cascade. He determines that the members of the crowd know or understand something he doesn't and that he will gain financially by adopting the crowd's investment theme. This crowd cascade can continue so long as the crowd's successful investment stance persuades new investors to join. But remember that most members of the crowd have voluntarily chosen to ignore their own private information as they cascaded into the crowd. This is a ticking bomb that will inevitably blow the crowd apart. As soon as the investment performance of the crowd falters or convincing, contradictory information becomes available to crowd members, the cascade starts to run in reverse. At that point the crowd's members will no longer weight the crowd's beliefs more heavily than their own. The investment crowd disintegrates, vanis.h.i.+ng like a wisp of smoke in the wind.

Here we reach the critical point in our a.n.a.lysis of investment crowds. We know that a crowd built by an information cascade is a fragile one. Information that contradicts the crowd's theme can stop the cascade and cause the crowd to disintegrate. What kind of information might be most effective in reversing the cascade of new members into the crowd? I think the answer will be obvious after a little reflection. The crowd's growth is fed by the success of the crowd's investment theme. Without evidence that prices are moving in the direction predicted by the crowd's theme, the cascade would stop. Prospective crowd members would no longer weight the crowd's information more heavily than their own. The market's action itself can halt the cascade.

But for the market to move against the combined buying or selling power of an investment crowd, some sort of countervailing coalition of speculators and investors must develop. This can happen only if the market price has been driven far enough from fair value by the crowd's activities to make investing opposite the crowd's theme an attractive and reasonably safe proposition. It is important to note that it isn't necessary for yet another crowd to form in opposition to the first one. All that is required is that the divergence between price and fair value created by the crowd's activities attracts enough value investors with an opposite view that the movement of price away from fair value is halted.

Because an information cascade is so fragile, the growth of an investment crowd is likely to halt as soon as the above-average returns to its investment theme fail to materialize. This will happen as a natural consequence of the significant divergence of the market price from fair value that has resulted from the crowd's investment activities. And as soon as the crowd's growth stops, there will then be a trickle of members who lose faith in the crowd's theme. As they leave the crowd, the market price will slowly begin to drift back toward fair value. At that point the information cascade that built the crowd will begin to run in reverse and the trickle of disillusioned members will become a flood.

A market affected by information cascades and the investment crowds cascades build will experience many episodes of significant overvaluation and undervaluation. The fact that market prices fluctuate so much more than can be explained by long-term valuation factors tells us one thing: Cascade-induced valuation mistakes are the rule, not the exception, in financial markets. These mistakes are investment opportunities for a contrarian trader.

How can the contrarian trader recognize an information cascade in the making? What clues should he look for to detect an emerging investment crowd and the investment opportunities it creates? In the next chapter we start to answer these questions by building a detailed description of the life cycle of a typical investment crowd.

CHAPTER 5.

The Life Cycle and Psychology of an Investment Crowd The cycle of birth and death * * cosmological a.n.a.logy cosmological a.n.a.logy * * business compet.i.tors and value investors eventually cause the death of the crowd business compet.i.tors and value investors eventually cause the death of the crowd * * once the crowd disintegrates, a new crowd often starts to form in response to the extended movement of prices once the crowd disintegrates, a new crowd often starts to form in response to the extended movement of prices * * crowd formation causes (and in turn is caused by) excessive price volatility crowd formation causes (and in turn is caused by) excessive price volatility * * bearish crowds are different from bullish ones bearish crowds are different from bullish ones * * the 1994-2000 stock market bubble the 1994-2000 stock market bubble * * stock market valuation and Tobin's q ratio stock market valuation and Tobin's q ratio * * it's different this time it's different this time * * the new information economy the new information economy * * shattered dreams shattered dreams * * the bear crowd of 2000-2002 the bear crowd of 2000-2002 * * the quest for certainty the quest for certainty * * the conflict between science and certainty the conflict between science and certainty * * every opinion has its rationale every opinion has its rationale * * instinctual belief instinctual belief * * the need for affirmation the need for affirmation * * pied pipers lead the crowd pied pipers lead the crowd * * mental unity of crowds mental unity of crowds * * intolerance of contrary views intolerance of contrary views * * examples from the 1994-2000 bubble examples from the 1994-2000 bubble * * Julian Robertson, Stanley Druckenmiller, and Gail Dudack Julian Robertson, Stanley Druckenmiller, and Gail Dudack * * Allan Sloan and America Online (AOL) Allan Sloan and America Online (AOL) * * social and financial pressure on unbelievers social and financial pressure on unbelievers * * price volatility and h.o.m.ogeneous thinking in crowds price volatility and h.o.m.ogeneous thinking in crowds * * price volatility is one sign that a crowd is mature price volatility is one sign that a crowd is mature

PROLOGUE.

In this chapter I collect the strands of fact and theory we've developed in the preceding chapters. I will weave them into a tapestry that chronicles the life and death of a typical investment crowd. Investment crowds differ from one another in numerous details, but they all develop through the stages depicted on this tapestry. To make things as concrete as possible, I am going to use the stock market crowds of 1994-2002 to ill.u.s.trate the typical pattern of an investment crowd's life and death. Once this cyclical picture of the life of an investment crowd is completed, we will turn to a study of the internal life of an investment crowd. By this I mean the study of the psychological att.i.tudes of those who join the crowd and the nature of the individual rationalizations that permit the information cascade to develop. Knowledge of these individual patterns of thinking can be very valuable for identifying information cascades and the investment crowds that grow from them.

THE CYCLE OF BIRTH AND DEATH.

How do investment crowds get started? There is no single right answer to this question. But I think it is accurate to say that most investment crowds find their genesis in the deaths of other investment crowds. I like to use a very apt cosmological metaphor to help understand this process. Investment crowds are the stars of the financial universe. The stars in the Milky Way and in the much larger cosmos have limited lifetimes, which typically end in a ma.s.sive explosion called a supernova. But new stars are being born (i.e., starting their own process of nuclear fusion) all the time. What is the source of the material that is the stuff of a new star? Well, it is just the cosmic debris left by the explosions of old stars!

In much the same way, investment crowds burn brightly in the financial universe and are responsible for much of the observed price fluctuation. But they have finite lifetimes (a few months to a few years). The inevitable disintegration of any investment crowd causes a big run-up or drop in prices and lots of commotion and confusion in the marketplace. But the debris a.s.sociated with the disintegration of a crowd is the stuff from which the next crowd forms. The change in price a.s.sociated with the disintegration of an investment crowd is powerful advertising. It attracts the attention of investors, especially of those whose portfolios have been directly affected by the rise or fall in a.s.set value.

Let's examine this process more carefully. To keep things concrete, you might want to keep in mind as an example the price of a common stock or of some market average.

Imagine that a bullish investment crowd has driven the average or the stock price well above fair value. Every investment crowd's lifetime is limited by the natural economic forces of supply, demand, and compet.i.tion. Once price is far enough above fair value, willing sellers appear as if out of the woodwork. New business compet.i.tors arise, and there may be a capital investment boom in the relevant industries. This s.h.i.+ft in supply conditions steadies the price above fair value and eventually causes it to drop. The fragility of every information cascade guarantees that this process will feed on itself, eventually leading to the complete disintegration of the bullish crowd.

But this dramatic drop back toward fair value attracts the attention of another group of investors. The fear of financial loss unites them into a bearish crowd and leads to the articulation of their bearish investment theme. Their activities force price well below fair value. Typically this is a very temporary situation. Price excursions below fair value are generally shorter in duration than those above fair value. Once the fear of financial ruin dissipates, prices return fairly quickly to fair value. The movement is reinforced by the natural economic forces if they have had time to operate and if the bearish market mistake is large enough. The return to fair value accompanies the disintegration of the bearish information cascade and sets the stage for the birth of a new, bullish investment crowd.

Now, the next bullish investment crowd does not form when price is below fair value. Instead the rise in price from far below fair value back toward fair value advertises a bullish environment. It is this price movement that stimulates the birth of a bullish crowd. Some investors have hit the jackpot because of this run-up in prices. If this jackpot is big enough to attract the attention of the media and of other investors, a bullish investment crowd will form. The bullish crowd is most likely to form at the point where price has returned to but is not yet above fair value.

We have before us a cyclical picture of the self-reinforcing life cycles of investment crowds. This picture offers an explanation of the connection between investment crowds and excessive stock market fluctuations and volatility. Bullish and bearish investment crowds in turn drive the market price above and then below fair value. Each such swing sets up the conditions that eventually lead to the formation of another crowd with the opposite market orientation.

So far I have depicted bullish and bearish investment crowds pretty much as opposite sides of the same coin. It is tempting to believe that there is a pleasing symmetry that controls the way both sorts of crowds form, grow, and dissolve. But as anyone with experience in the financial markets will testify, this isn't true. In fact, the growth of a bullish crowd generally proceeds at a more leisurely pace than does the growth of a bearish crowd. Bullish crowds last longer, and the mistakes they force upon markets tend to extend over longer periods of time. In contrast, bearish crowds form and dissolve over relatively short time spans, and the market mistakes they cause are for the most part of comparatively short duration. The reasons for this asymmetry are not very well understood. It may arise from the fact that there is no specific limit to how high a price may climb, but no price can drop below zero. In any case, we will find that this difference between bullish and bearish crowds is itself evidence pointing to the essential characteristic of all investment crowds: Their members eventually behave as a herd-a ma.s.s of individuals whose behavior is governed by instinct, suggestibility, and imitation, not by reason.

THE STOCK MARKET BUBBLE OF 1994-2000.

Bullish investment crowds get started when the market is trading near fair value. The birth of the crowd is triggered by the substantial advance in prices that occurred as the market returned to fair value from a position well below fair value. The phenomenon is nicely ill.u.s.trated by the birth of the investment crowd that gave us the stock market bubble of 1994-2000.

The stock market boom of 1994-2000 in the United States was the culmination of an 18-year advance, an unprecedented bull market that began in 1982. At its low in 1982 the Dow Jones Industrial Average stood at 777. On the final trading day of 1994 the average closed at 3,835, almost 400 percent higher than in 1982. As the market continued its advance in 1995, many experts believed that the averages were already trading well above fair value and on this basis predicted an imminent crash. But were prices really really above fair value then? above fair value then?

As every investor knows, a.s.sessments like this are hard to make. But I have found that, at least as far as long-term trends in the U.S. stock market are concerned, it is very helpful to seek guidance from the famous q ratio q ratio, which was invented by the n.o.bel Prize-winning economist James Tobin. Tobin believed that to determine the extent to which the stock market is under- or overvalued, it makes sense to compare the value the stock market places on corporate a.s.sets with the current cost of replacing these a.s.sets (their so-called replacement value). It is this ratio of stock market value to replacement value that Tobin christened q. q.

The best explanation of the q ratio I have found in print appears in the book Valuing Wall Street Valuing Wall Street by Andrew Smithers and Stephen Wright (McGraw-Hill, 2000). In it they explain the q ratio and apply it to practical valuation problems. A simple synopsis of the theory behind the q ratio goes like this. When q is substantially above 1.0 it is cheaper to buy real a.s.sets, build factories, buy equipment, and start businesses than it is to buy the same stream of income in the stock market. Thus a q above 1.0 stimulates a boom in real economic investment and this leads to above-average economic growth. A q below 1.0 means that it is cheaper to buy a given stream of income in the stock market than it is to earn it by making real investments in the economy. Thus a q below 1.0 acts as a brake on the economy, or at least leads to below-average economic growth. by Andrew Smithers and Stephen Wright (McGraw-Hill, 2000). In it they explain the q ratio and apply it to practical valuation problems. A simple synopsis of the theory behind the q ratio goes like this. When q is substantially above 1.0 it is cheaper to buy real a.s.sets, build factories, buy equipment, and start businesses than it is to buy the same stream of income in the stock market. Thus a q above 1.0 stimulates a boom in real economic investment and this leads to above-average economic growth. A q below 1.0 means that it is cheaper to buy a given stream of income in the stock market than it is to earn it by making real investments in the economy. Thus a q below 1.0 acts as a brake on the economy, or at least leads to below-average economic growth.

Thus one expects a q ratio substantially above 1.0 to accompany an economy-wide investment boom, which arises because the stock market is overvalued. In such a situation investors want to buy capital equipment and form new corporations, then sell their interest in the stock market which values this interest well above its cost. In contrast, a q ratio well below 1.0 should be a.s.sociated with weak demand for new capital goods. The stock market is substantially undervalued and making real investments yields an instantaneous stock market loss since the market values them at less than cost.

Here is a remarkable fact: At the end of 1994, after a nearly 400 percent rise in stock prices measured from the 1982 low point, Tobin's q ratio was just barely above 1.0! Over the preceding 110 years q had varied from a low value of 0.4 to a high value of 1.9. In 1994, q stood at 1.1. By contrast, in 1982 q stood at 0.5, the lowest value seen since 1932 in the depths of the Great Depression. So we see that, at least in terms of Tobin's q ratio, a 12-year, 400 percent advance in stock prices had only returned the market to fair value by the end of 1994!

This long advance had made a big impression on investors. To see how it changed people's thinking about the stock market, I highly recommend reading Bull! Bull! (HarperCollins, 2004), a book by Maggie Mahar that chronicles the stock market boom and bust from 1982 through 2002 in fascinating detail. (HarperCollins, 2004), a book by Maggie Mahar that chronicles the stock market boom and bust from 1982 through 2002 in fascinating detail.

Mahar reports that in 1995, for the first time since the early 1970s, U.S. households held more wealth in the stock market than in real estate. I date the emergence of the bull market investment crowd from 1995. This crowd's theme was that buying shares in mutual funds and holding them was the sure way to ama.s.s wealth and achieve early retirement. The universal expectation of this crowd was that stocks would return 10 to 20 percent yearly in perpetuity. During the next five years money poured into mutual funds. People who had never invested in stocks committed their life savings to the market. The q ratio rose from 1.1 in 1994 to an unprecedented peak of 2.6 in the year 2000. The 2.6 reading was the highest q had ever been in the 120 years for which data had been available! It is not then surprising that the 1998-2000 period saw an enormous capital goods boom, which vastly increased capacity in the telecom, computer equipment, and Internet services industries.

The birth of this investment crowd in 1995 occurred when the stock market was priced at fair value, but only after a long rise in prices over the preceding 12 years from a level well below fair value. This very powerful stimulus no doubt played a role in determining this crowd's unusual longevity and the size of the mistake it forced upon the market. That aside, this is a cla.s.sic ill.u.s.tration of the circ.u.mstances attending the birth of most investment crowds. But investment crowds can develop in other ways as well, and such crowds also played an important role in the 1994-2002 boom and bust.

IT'S DIFFERENT THIS TIME: THE NEW INFORMATION ECONOMY In Chapter 4 we saw that an information cascade begins when some investors decide that other investors know more about a particular investment opportunity than they do themselves. In such circ.u.mstances it can be rational for a person to imitate another's actions even when his own private information and proclivities tug him in the opposite direction.

So it is natural to expect that an information cascade is especially likely to develop in response to a genuinely new and different investment opportunity, one that is completely outside the realm of most investors' personal experience. And this is exactly what is happening when we hear talk about new industries and new technologies that promise to revolutionize the economy. The unfamiliarity of the new investment opportunity can make it difficult to tell if the market price of an initial public offering or of the stock of a company in an unfamiliar industry is near fair value. Such circ.u.mstances make it easy for business and investment cheerleaders to a.s.sert that the price is at or even below fair value and thus to encourage the growth of an investment crowd.

The 1994-2000 stock market boom was accompanied by a popularization of the phrase the new economy. new economy. This referred to the confluence of trends toward This referred to the confluence of trends toward globalization globalization on the one hand and the growing use of on the one hand and the growing use of information technology information technology (embedded in communication and computer equipment) to revolutionize the way business is done. In the new economy, productivity growth was thought to be unusually high. This permitted high economic growth and high employment accompanied by low inflation, a situation that Goldilocks herself would envy. (embedded in communication and computer equipment) to revolutionize the way business is done. In the new economy, productivity growth was thought to be unusually high. This permitted high economic growth and high employment accompanied by low inflation, a situation that Goldilocks herself would envy.

The new economy environment of 1994-2000 not only encouraged the growth of the mutual fund stock market crowd but also supported the emergence of smaller investment crowds, which focused on particular industries or companies. Since information technology was supposedly revolutionizing economic relations.h.i.+ps, old thinking about the relations.h.i.+p of fair value to earnings potential became suspect. These newcomers to the corporate world understood things that old-time investors did not. Conditions were ripe for the development of information cascades, and develop they did. Soon investment crowds took hold in the stocks of companies like Amazon, America Online (AOL), Apple Computer, Dell Computer, eBay, Enron, Intel, Lucent Technologies, MCI WorldCom, Microsoft, Oracle, Priceline.com, Qwest Communications, Silicon Graphics, and Yahoo!, to name just a few. These telecommunications and dot-com companies rocketed the NASDAQ-100 index to a gain of almost 400 percent from late 1998 to early 2000.

The very newness of the businesses these companies saw themselves in made it easy to convince normally skeptical investors that traditional valuation standards no longer applied. As the resulting information cascade gained momentum, the enormous amount of money made by entrepreneurs and early investors gave added credence to this theme. That it all ended badly is just the ever-repeating story of investment crowds and stock market information cascades.

SHATTERED DREAMS: THE BEAR CROWD OF 2001-2002.

So far we have looked at the birth of the bubble crowds that formed during the stock market run-up from fair value in 1994 to extreme overvaluation in 2000. These bullish investment crowds grew from cascaded dreams of a new economy. When these bubbles burst, the cascade of dreams morphed into a cascade of fear and recrimination. This built the bear market crowd that dominated the U.S. stock market during 2001-2002.

I have explained the process by which the typical bearish crowd develops. The disintegration of the preceding bullish crowds triggers the birth of new bearish crowds. This generally happens only after price has returned to the vicinity of fair value.

Remember that the fair value price is only a signpost on the road to the inevitable undervaluation that will be forced by the growth of the bearish investment crowd.

Where was fair value in 2001 for the U.S. stock market? At the peak of the bull market in 2000, Tobin's q ratio stood at a historical high of 2.6, exceeding by a very wide margin its previous high of 1.9. The ratio was clearly behaving differently than it had over the past 120 years by leaving its normal range of fluctuation between 0.4 and 1.9. This posed a problem for anyone trying to use q in real time, because it raised measurement reliability issues. Perhaps the numbers used to calculate q didn't have the same economic significance they had in the past?

These uncertainties forced me to temporarily abandon q as an estimator of fair value for the U.S. stock market in 2001-2002. The q ratio may yet prove to be of great use, but the jury is out on the current significance of its readings to investors. This sort of circ.u.mstance ill.u.s.trates the importance of having multiple methods for estimating fair value. When one or another method seems to be out of whack for some reason, one has alternative measurement available. Here is a simple alternative method I like for estimating long-term fair value. It uses only market price data, not economic data. As an empirical matter, the U.S. stock market has followed a cycle of about 48 months from trough to trough over the period that began in 1929. So as a rough-and-ready estimate of fair value, I like to use the 48-month, simple moving average of the month-end reading of the S&P 500 stock market index. (Recall that this moving average is computed by adding up 48 consecutive monthly closes and dividing the answer by 48.) The 48-month length of the moving average is selected to minimize the effects of the normal 48-month rhythm is stock prices.

Let's get back to the story of the 2001-2002 bearish investment crowd. In March 2001 the S&P 500 touched its 48-month moving average, my long-term estimate for fair value, for the first time since 1982. The index had traded above this fair value estimate for more than 18 years! The moving average stood at roughly 1,210 in March 2001, and the index itself dropped to 1,084 that same month and then rallied to 1,315 by May 2001.

The return to fair value from a point of overvaluation will leave investors with losses. Indeed, newspaper and magazine commentary during late 2000 and early 2001 held unanimously that the dot-com and technology bubbles had burst. This event was dramatically recorded by the NASDAQ Composite stock market index, home to most of the bubble stocks. At its 5,132 high in March 2000, this index was trading more that 150 percent above fair value as measured by its 48-month moving average. One year later the NASDAQ Composite had dropped as low as 1,619, about 35 percent below the moving average estimate of fair value at the time and a stomach-churning 69 percent drop from its high a year earlier.

The shock of a 69 percent drop in the technology sector was transmitted instantly to investors, because they had poured money into technology mutual funds during the 1999-2000 bubble years at unprecedented rates. The natural response to such destruction in portfolio values is fear of further loss, a search for means of escape, and the identification of scapegoats and evildoers to be held responsible for the damage. These are the potent ingredients from which the theme of nearly every bearish investment crowd emerges. Bearish crowds tend to develop quickly because most members already have experienced financial loss. The information cascade that develops on these foundations gives new members of the crowd reason to fear that more erosion in a.s.set values lies ahead. Such persuasion is easily accomplished. It relies on the universal convention that guides people in markets and in life: Tomorrow will be pretty much like today Tomorrow will be pretty much like today. But in the case of a bearish investment crowd, today is filled with financial pain. Tomorrow can only be worse, especially because the crooks responsible for the current distress have not all been caught. The means of escape becomes obvious: sell. Once an investor has sold out, he is even more convinced that worse is to come in the market, for otherwise his actions will appear foolish.

I will discuss the 2001-2002 bearish stock market crowd in more detail in subsequent chapters. Suffice to say here that by mid-2002 it had become a very visible and dominating investment crowd. By that time the NASDAQ Composite index was trading more than 50 percent below its moving average estimate of fair value and nearly 80 percent below its 2000 high. The S&P 500 was trading 38 percent below its fair value estimate, down about 50 percent from its 2000 high at 1,553.

POPULAR INSTINCTS AND THE SEARCH FOR CERTAINTY.

As a bullish crowd disintegrates, price drops back toward fair value. When a bearish crowd disintegrates, price rallies upward toward fair value. In both instances a big, unexpected price change advertises the death of the crowd. It also attracts the attention of the rest of the financial world. People ask, "Why did prices drop so much?" They wonder, "How come the market rallied so far in the face of bad news?"

Investors always want answers. In this respect they are simply being human. It is man's inclination to ask and to answer the question why why that distinguishes him from animals lower in the food chain. Scientists grapple with the that distinguishes him from animals lower in the food chain. Scientists grapple with the why why question every day. That's their business. But, as every scientist knows, the most powerful device humans have invented for getting answers, the scientific method, requires a suspension of judgment about this answer or that one. Instead the scientist proposes a theory to explain observed events and then carefully compares it with the empirical evidence. Sometimes the data prove the theory is plain wrong. Or, if he is lucky, the scientist might find that the data are entirely consistent with the theory's predictions. But these outcomes are not common. Usually the comparison of data with theoretical predictions has an ambiguous result, so the scientist must go back to his desk or lab to reconsider his ideas and research strategy. The s.h.i.+p of scientific inquiry spends little time near the sheltered sh.o.r.e of certainty. question every day. That's their business. But, as every scientist knows, the most powerful device humans have invented for getting answers, the scientific method, requires a suspension of judgment about this answer or that one. Instead the scientist proposes a theory to explain observed events and then carefully compares it with the empirical evidence. Sometimes the data prove the theory is plain wrong. Or, if he is lucky, the scientist might find that the data are entirely consistent with the theory's predictions. But these outcomes are not common. Usually the comparison of data with theoretical predictions has an ambiguous result, so the scientist must go back to his desk or lab to reconsider his ideas and research strategy. The s.h.i.+p of scientific inquiry spends little time near the sheltered sh.o.r.e of certainty.

In their everyday affairs people do not use the scientific method to explain events. They have neither the time nor the skill. In his book Instincts of the Herd in Peace and War Instincts of the Herd in Peace and War (orig. pub. 1919; reprinted by Cosimo Cla.s.sics, 2005), the London surgeon and sociologist Wilfred Trotter observes: "In matters that really interest him, man cannot support the suspense of judgment which science so often has to enjoin. He is too anxious to feel certain to have time to know." With little time to know (i.e., to apply the scientific method) and a low tolerance for ambiguity, the typical investor must rely more on instinct than on science to explain market movements. (orig. pub. 1919; reprinted by Cosimo Cla.s.sics, 2005), the London surgeon and sociologist Wilfred Trotter observes: "In matters that really interest him, man cannot support the suspense of judgment which science so often has to enjoin. He is too anxious to feel certain to have time to know." With little time to know (i.e., to apply the scientific method) and a low tolerance for ambiguity, the typical investor must rely more on instinct than on science to explain market movements.

What do I mean by the word instinct instinct? Nowadays psychologists and sociologists don't believe that any human behavior is instinctual in a narrow technical sense, that is, in the same sense that certain animal behavior may be described as instinctual and biologically based. But I think this term can still be useful in describing the nature of rationalizations people use to explain their beliefs and actions. This is precisely the way Trotter uses it in his book. He cites Volume 2 of William James' Principles of Psychology Principles of Psychology (1890, reprinted by Dover Books in 1950 and available digitally on the Internet) to explain how instinctual behavior in Trotter's sense appears to human introspection. He defines an instinctive action or belief as one appearing so grounded in common sense that any idea of discussing its basis appears "foolish or wicked." An instinctual action is obviously the right thing to do. Only a foolish or ill-intentioned person would dare dispute an instinctual belief. (1890, reprinted by Dover Books in 1950 and available digitally on the Internet) to explain how instinctual behavior in Trotter's sense appears to human introspection. He defines an instinctive action or belief as one appearing so grounded in common sense that any idea of discussing its basis appears "foolish or wicked." An instinctual action is obviously the right thing to do. Only a foolish or ill-intentioned person would dare dispute an instinctual belief.

Of course instinct in this sense is not a biological phenomenon and so is unlike animal instinct. Instead it is a.n.a.logous to a religious belief or a belief that is axiomatic, based on a.s.sumptions that are never questioned. Instinctual beliefs in Trotter's sense often arise when people transfer their experiences and behavioral responses and beliefs from one realm of activity to another. But an even more important mechanism for belief transmission and acquisition is the social group and the crowd.

People prefer the comfort and certainty of instinctive belief to the ambiguity a.s.sociated with scientific procedure and knowledge. Trotter points out that in everyday life: [individuals] make vast numbers of judgments of a very precise kind upon subjects of very great variety, complexity, and difficulty. . . . The bulk of such opinions must necessarily be without rational basis . . . since they are concerned with problems admitted by the expert to be still unsolved, while as to the rest it is clear that the training and experience of no average man can qualify him to have any opinion upon them at all.

Trotter continues: It is clear at the outset that these beliefs are regarded by the holder as rational, and defended as such, while the position of one who holds contrary views is held to be obviously unreasonable.

To be sure, no one I have ever met has admitted that his beliefs are for the most part instinctual in this sense. Far from it. We all think our beliefs are rational, often self-evidently so. Trotter notes this phenomenon: [I]t should be observed that the mind rarely leaves uncriticized [its] a.s.sumptions . . . the tendency being for it to find more or less elaborately rationalized justification for them. This is in accordance with the enormously exaggerated weight ascribed to reason in the formation of opinion and conduct.

If most of our beliefs are instinctual, it is natural to wonder about their source. How do such beliefs arise? How do we acquire them? Certainly some are generalizations from individual experience. But we all hold beliefs about politics, national affairs, economics, local affairs, sports, and so forth, and these realms are for the most part outside of our personal experience.

Most of our beliefs and their rationalizations are adopted from the social groups in which we partic.i.p.ate. We take the very fact that many people whom we know hold these beliefs to be evidence for their validity. This mechanism for belief formation has something of the flavor of an information cascade. But to describe it in cascade terms ascribes too much rationality to the process of belief acquisition. Trotter expresses this same idea more colorfully when he compares society to a herd of animals, but a herd endowed with a voice, a power of suggestion: [B]elief in affirmations . . . sanctioned by the herd is a normal mechanism of the human mind, and goes on however much such affirmations may be opposed by evidence. . . . [R]eason cannot enforce belief against herd suggestion. . . . [T]otally false opinions may appear . . . to possess all the characters of rationally verifiable truth.

This is not to say that all or even most of our beliefs that have a purely social basis are necessarily wrong or irrational. Indeed, scientifically verifiable beliefs often acquire the sanction of the herd, the crowd, or the social group and are thus transmitted to those who have no access to or understanding of the scientific method. This sanction generally takes a generation or more to acquire, but one need go no further than the examples of Darwin's theory of evolution or Einstein's relativity theory to see this process in action on a large scale.

THE PIED PIPERS OF INVESTMENT CROWDS.

The drop in the market price back to fair value during the death of a bullish crowd and the rise in price back to fair value during the death of a bearish crowd generally attract a lot of attention. Investors want an explanation of a dramatic and unexpected price change. This demand for explanation naturally creates its own supply. I liken the rationalizations that emerge in these situations to the Pied Piper of Hamelin. The rationalizations rarely if ever have any scientific basis. Instead, the music of the pipe first attracts attention because it sounds like a plausible, indeed logical, explanation of an otherwise mysterious price movement. The reasoning appears instinctually obvious to its growing number of adherents. But eventually the piper exacts a toll from his audience, a price that far exceeds the few coins rational calculation would have offered. The investment crowd follows the piper to its inevitable doom brought about by the economic forces that limit the size of market mistakes.

During the big move in crude oil prices from $40 in 2004 to $140 in 2008, the pied piper of peak oil played a very seductive tune. Peak oil advocates a.s.serted that world production of crude oil was bound to start an inevitable decline for purely geological reasons early in the twenty-first century. The obvious conclusion was that crude oil prices had nowhere to go but up. Other theories competing with peak oil reinforced expectations of a steady upward march in crude oil prices. One of these was that the accelerated economic growth in China and India was creating unusual and expanding demand for all natural resources, not just for oil. Another was the devil theory devil theory of markets, which attributes all big moves to the machinations of speculators. Here were three pied pipers that enabled the bullish crude oil crowd to grow to enormous size by mid-2008. of markets, which attributes all big moves to the machinations of speculators. Here were three pied pipers that enabled the bullish crude oil crowd to grow to enormous size by mid-2008.

The new economy bubble of 1994-2000 had its own pied pipers. This bubble inflated on the basis of the new information economy, the acceleration of globalization, and the a.s.sociated rise in productivity. Demand for telecom bandwidth was said to be unlimited. The Internet was to provide a whole new business model in which profits played but a small role. One interesting aspect of the bubble was that many of the pied pipers were actual people. They had names everyone in the bubble crowds recognized: Jack Grubman, Frank Quattrone, Abby Joseph Cohen, Mary Meeker, Henry Blodget, Maria Bartiromo, Ralph Acampora, and Alan Greenspan.

The bear crowd of 2001-2002 had its own pied pipers, but they were not as plainly visible as the ones that had enabled the preceding bubble. The bear crowds that form amidst the debris of a bubble typically blend the characteristics of lynch mobs and repentant sinners: We had it coming for believing in the patent nonsense of the bubble investment themes; they led us astray and cheated us, and now they must pay the price. The crowd searches for scapegoats, and those it finds become inverted pied pipers, repelling listeners with the dissonant music of their pipes. They are offered as exemplars of unsound and even stupid investment policies. This process of tearing down the edifice of belief that supported the bubble continues until only discouragement and fear for the future control investor att.i.tudes.

THE MENTAL UNITY OF INVESTMENT CROWDS.

So far we have seen that investment crowds get started after the market price returns to fair value following a significant excursion away from fair value. The price movement that arises from the return to fair value attracts public attention. Investors' human aversion to ambiguity and uncertainty, coupled with their limited capacity for scientific thinking, leaves them vulnerable to pied piper explanations for the run-up or the drop in prices. These explanations are always plausible and exhibit a certain internal logic. This is what makes them seductive tunes for the pied piper's pipe. If the tune promises a significant continuation of the price move that has already been observed, then an investment crowd will be born and start to grow.

Now we will discuss the view from inside a developing investment crowd. How do people in a crowd act? How can we distinguish an investment crowd from any random group of individuals? I am going to offer my own answers to these questions, but much of what I say can also be found in Gustav Le Bon's 1895 cla.s.sic The Crowd. The Crowd. The a.n.a.lysis of crowd motivation and behavior you can find in his book makes it required reading for any contrarian trader. The a.n.a.lysis of crowd motivation and behavior you can find in his book makes it required reading for any contrarian trader.

The most important characteristic of any crowd is what Gustav Le Bon calls the crowd's mental unity mental unity. The attention and sentiments of all crowd members are focused in a single direction or on a single phenomenon. The mental unity of an investment crowd is externally manifested in the crowd's investment theme investment theme. An investment theme is a basket of explanations and of the forecasts that seem to be their obvious consequences. An investment theme identifies an a.s.set or a cla.s.s of a.s.sets, and explains why the price of these a.s.sets has changed so much recently (during the return to fair value) and why price will continue to change in the same direction. The pied pipers of the financial world are the early advocates of these investment themes. We have already seen examples of investment themes: those that unified the bubble crowds of 1994-2000 and the bear market crowd of 2001-2002.

An investment theme begins to attract adherents as its true believers publicize their views and as the market price continues to move in the direction they forecast. This is the beginning of an information cascade. Newcomers are persuaded by the arguments of the true believers to ignore their own information (if they have any) and accept the information upon which the cascade builds. The mental unity of the crowd begins its development in this way. But there is much more to a crowd's mental unity than simply its investment theme.

The essence of life within a crowd is constant reinforcement and affirmation of the crowd's beliefs. Crowd members communicate with one another, either directly as individuals or indirectly through the print and electronic media. Crowd leaders are always visible and always courting publicity. They never forgo the chance to "talk their book," to urge newcomers to accept the crowd's investment theme.

This constant repet.i.tion of a crowd's investment theme is the single most important characteristic of life within an investment crowd. Life within any crowd (not just an investment crowd) is like life within an echo chamber. All one hears is a repet.i.tion of the investment theme and confident p.r.o.nouncements about profits to be earned by its followers. Soon even the logic supporting the theme is forgotten-or reduced to easily remember cliche. The mere fact that crowd members experience so much affirmation of their belief (albeit only from other crowd members) is taken as more evidence of the correctness of the crowd's theme.

Both Le Bon and Trotter observe that the affirmative messages that build crowd unity are not usually appeals to the intellect of crowd members. Instead they are appeals to emotion, to stereotypes, to dreams or fears. The language of persuasion and crowd solidarity is the language of drama, not science.

Affirmation and repet.i.tion build the mental unity of an investment crowd. As the crowd's mental unity develops, members of the crowd begin to resemble one another in their preferences, beliefs, and actions. Crowd members sense an increase in the power of the crowd, and this encourages each of them to be more of a risk taker. Of course what an outsider would identify as a riskier investment stance the crowd members see as very safe, almost a sure thing. Crowd members give little thought to the consequences of being wrong. Such a discussion would not be sanctioned by the crowd, and doubters would be cast to the crowd's fringes or expelled.

I have found a crowd's intolerance of contrary views to be its single most important identifying characteristic. This intolerance shows itself in the form of ridicule and abuse of any skeptical consideration of its theme. I have written an investment blog (www.carlfutia.blogspot.com) for several years. I try to take the viewpoint of a contrarian trader. When the comments on my blog are the most numerous and abusive I know that a market turn is imminent.

Crowd intolerance of contrary views can manifest itself in other, more dramatic ways in the world of a.s.set management by mutual funds, pension funds, and hedge funds. Money managers who do not join the crowd run a significant risk of being fired or put out of business. The same fate can befall any investment guru who dares cross the crowd. Both these sorts of things happened in dramatic fas.h.i.+on during the late stages of the 1994-2000 stock market bubble.

Founder Julian Robertson closed the investment firm Tiger Management in March 2000, right at the top of the stock market bubble. He did extraordinarily well for his investors, earning an average annual return net of fees of 25 percent during 20 years. But the bubble year of 1999 was his undoing as his fund dropped 19 percent while the S&P 500 rose 21 percent. His investment method couldn't account for the bubble crowd's ma.s.sive buying power. His poor performance in 1999 would have caused his investors to abandon him in droves had he not shut down his firm when he did. So one of the most talented hedge fund managers of his generation was forced out of business by the actions of the bubble crowd, whom he had chosen to ignore.

If you have a taste for irony, you will find the saga of Stanley Druckenmiller even more intriguing. Druckenmiller ran a good part of George Soros's hedge fund, the Quantum Fund, from 1988 through 2000. He was and is an enormously talented investor. But like so many experienced investors, he did not partic.i.p.ate in the late 1990s technology boom, believing that it would come to no good end. This reluctance began to hurt the Quantum Fund as its investors questioned the wisdom of staying out of the technology sector.

During the summer of 1999 Druckenmiller had an epiphany about the tech sector while he was attending an investment conference. Upon his return he put the Quantum Fund into many of the technology stocks that were rising dramatically during the latter half of 1999, the terminal stage of the NASDAQ Composite index's 400 percent rally from its 1998 low point. The Quantum Fund's performance during the latter half of 1999 was outstanding. Sadly, Druckenmiller's tech portfolio was buried by the dramatic sell-off during March-May 2000, which was eventually to bring the NASDAQ down 80 percent from its 2000 high. This setback led to Druckenmiller's departure from the Quantum Fund and to Soros's temporary shutdown of the fund's speculative activity.

The stories of these two talented investors, Robertson and Druckenmiller, ill.u.s.trate well the fate of money managers who don't align themselves with dominant investment crowds. The crowd will abandon even the most talented manager if he is not committed to the crowd's investment theme. Every manager knows this, and all understand the risk of being put out of business for failing to toe the party line in their portfolios. This of course reinforces the influence of any investment crowd. The irony of Stanley Druckenmiller's experience is that he bent to the crowd's pressure at precisely the wrong time. The resulting portfolio whipsaw severed his relations.h.i.+p with the Quantum Fund.

The story of Gail Dudack and the TV show Wall $treet Week Wall $treet Week also ill.u.s.trates the power of an investment crowd to harm the careers of those who oppose it. Dudack was chief market strategist during the 1990s at the brokerage firm S.G. Warburg. Like so many experienced market watchers, Dudack recognized the stock market bubble for what it was, and in late 1997 she announced her bearish views to her clients. She was also a regular guest on also ill.u.s.trates the power of an investment crowd to harm the careers of those who oppose it. Dudack was chief market strategist during the 1990s at the brokerage firm S.G. Warburg. Like so many experienced market watchers, Dudack recognized the stock market bubble for what it was, and in late 1997 she announced her bearish views to her clients. She was also a regular guest on Wall $treet Week with Louis Rukeyser Wall $treet Week with Louis Rukeyser, a very successful weekly TV stock market show produced for the public broadcasting affiliates. By November 1999 her bearish views had so irritated Louis Rukeyser, the show's popular host, that he removed Dudack from his list of regular guests and from her slot in his Elves Index, a joint market a.s.sessment of the technical a.n.a.lysts who appeared on his show.

Rukeyser was in the business of attracting viewers, and as everyone in the entertainment business agrees, the way to do this is to give people what they want. By 1999 the bubble crowds had grown so intolerant of contrary views that Rukeyser felt compelled to throw Dudack under the bus.

Here we have before us the experiences of two prominent hedge fund investors and one prominent market strategist. All three fell before the intolerant onslaught of the bubble crowd during the brief period from November 1999 through May 2000. It is not a coincidence that all the stock market averages reached their highs in January-March 2000.

It is not surprising that the intolerance of investment crowds can affect financial journalists as well. In 1995 and 1996 Allan Sloan, a financial columnist at Newsweek Newsweek, wrote several columns expressing his skepticism about the accounting methods of America Online (AOL), a company that pioneered the commercialization of Internet connectivity. AOL's stock had soared since its initial public offering in 1992, and Sloan believed it to be grossly overvalued. Needless to say, his skeptical views were not welcomed by the investment crowd, who believed in AOL. He eventually stopped writing about AOL, which, with exquisite timing, reached a merger agreement with Time Warner in the spring of 2000. Why did he stop? Maggie Mahar quotes Sloan in her fascinating book Bull! Bull! (HarperCollins, 2004), which chronicles the bubble: (HarperCollins, 2004), which chronicles the bubble: I knew I was right, but whenever I published one of those stories, everyone would carry on. . . . Finally I just gave up. I shouldn't have, but I did. There are only a limited number of swings you can take-eventually you look like a crank.

The social and financial pressure an investment crowd can exert on disbelievers cannot be overestimated. While much of a crowd's growth occurs as the result of an information cascade, the continuous strengthening of a crowd's mental unity through constant affirmation, repet.i.tion, and dramatic price changes in the crowd's a.s.set makes this unity a powerful tool of persuasion. It magnifies the crowd's impact and importance and puts pressure on even the skeptics to join.

SUGGESTIBILITY, VOLATILITY, AND DISINTEGRATION.

As the mental unity of an investment crowd grows adherents to its investment theme become h.o.m.ogeneous in their thinking. When this happens the crowd develops an important characteristic that mature crowds share with herds of animals. Its members become very suggestible, and are apt to take action or modify their beliefs when presented with a strong image of something they desire or fear. Suggestibility opposes logical persuasion, but, as we have seen, every crowd member so strongly believes in the rationalizations offered by the crowd that no logical persuasion is necessary. Consequently, images presented to the crowd by its leaders or the media will be acted upon immediately by members of the crowd without further questioning.