PART II: CONTRARIANISM & STOCK MARKET IRRATIONALITY

At the end of Robert Prechter's "Elliott Wave Theorist" investment letter, the following explanation appears: "The Elliott Wave principle is a detailed description of how markets behave. The description reveals that mass investor psychology swings from pessimism to optimism and back in a natural sequence, creating specific patterns in price movement. (1)" The reason that Prechter believes market waves are caused by bipolar swings in popular mood is because at market tops and bottoms the consensus beliefs and expectations of investors are irrationally optimistic or pessimistic, respectively. Ups and downs in market prices are thus caused by the back and forth upset of investor expectations.

Because of stock market irrationality, a "contrarian" investment strategy has been developed that is today being used by some investors to "beat the market". More specifically, history shows that, since the investment crowd systematically errs, those who bet against the crowd will consistently gain higher investment returns than would be achieved by buying and holding. In order to show how this is so, below is an examination of studies and evidence that establish stock market irrationality and the advantage of going against popular opinion.

-The Empirical Case for Contrarianism-

Throughout the history of markets there has been recurring swings between mania and depression. Whether in pricing individual companies, or industries, or a nation's wealth, or the assets of the entire capitalist world, speculative manias push market prices to irrational highs and then panic and depression send prices to extreme lows. David Dreman, in his classic text The New Contrarian Investment Strategy, describes the pattern this way:

In the table below, Dreman showed how, even though separated by centuries, the effect on prices of extreme swings between mania and depression in mass psychology is always similar. Prices are first pulled up exponentially to nonsensical heights, and then prices come crashing down to equally absurd lows- usually to around one tenth to one twentieth of the previous price peak.

What is remarkable is that extreme, irrational mood swings in mass psychology are not exceptions of market history, they are the rule. While such absurd speculative manias as the Tulipomania, the South Sea Bubble, and the Roaring Twenties' stock market may be eye-catching, such insanity is intrinsic to market psychology all the time. The historical evidence demonstrates without question that market participants are radically subjective and systematically irrational.

As should be apparent, the implication of evidence showing market irrationality is that reality is contrary to what has been hypothesized by economists. As was explained in Part I, the Efficient Market Hypothesis which dominates contemporary economic thought is based upon the "rational expectations" assumption. For the sake of review, according to this assumption, investors, in making investment decisions, use all available, relevant information and the best known theories to form rational expectations of the future. Furthermore, investors reflect on past errors and adjust their information accordingly such that any errors are not repeated. The implications of the EMH and rational expectations assumption is that prices should reflect intrinsic value- i.e., no "over"- or "under"-valued assets exist, and price movements should be random and caused by unpredictable shocks. In reality, however, one finds that markets, especially the stock market, behave in the manner described at the end of Robert Prechter's "Elliott Wave Theorist" investment letter (see above). Investors, rather than forming rational expectations of the future such that prices tend toward intrinsic values and follow a random path, repeatedly form irrationally optimistic and then pessimistic expectations such that prices swing above and below intrinsic values and follow a predictable, cyclical path. On Wall street these cycles are commonly recognized as swings between mass greed and fear, something truly highlighted by the bubbles mentioned above.

In examining the evidence that shows how markets swing between irrational extremes of optimism and pessimism, greed and fear, let's start at the top. More specifically, since "expert" opinions and expectations help shape, or at least reinforce, popular opinions and expectations, let's examine the analytical and forecasting records of economists and business analysts.

-The Inaccuracy of Economic and Business Analyses-

"Stocks are now at what looks like a permanently high plateau."
(uttered by economist Irving Fischer a few days before Black Tuesday, 1929)

While no conclusive studies are available, there is substantial evidence that the forecasts of economists are consistently inaccurate. For instance, in 1947, a group of top economists predicted that U.S. economic activity would decline approximately 6 percent. That year the economy actually proved to be one of the strongest on record, showing an increase of 11 percent (4). A survey of many dozen economists and business analysts in late 1969, with the economy already in a downturn, disclosed that few believed a recession would occur in 1970. Contrary to expectations, there was a recession that year (5). In 1973, a survey of 32 major economic forecasters found that only one had projected a recession for the following year, 1974, which saw the worst recession of the post-war period up until that time (6). In 1981, supply-side "Reaganonomics" ruled the day and economists predicted booming growth and a shrinking budget deficit. Instead, there followed a booming deficit and shrinking economy (7). In early 1983, economists predicted that the emerging economic recovery would be below par at best. Thus began one the longest and strongest recoveries in post-war history (8).

Economists' failures at predicting the course of the economy is outweighed by an even worse record at forecasting inflation. In a study done by Geoffrey Moore on the forecasting record of the Council of Economic Advisors between 1962 and 1973 (8 years of which were characterized by relatively steady growth and inflation rates), Moore found that the Council's inflation forecasts correctly predicted the direction of change in the inflation rate in only 4 of the 13 years (9).

Incredibly, economists could improve the accuracy of their inflation forecasts by seeking the advice of households. In 1987, Michael Bryan and William Gavin of the Federal Reserve Bank of Cleveland compared the post war forecasting record of a semiannual survey of some 50 economists with the University of Michigan's consumer survey of 1,000 households. Even though the economists had sophisticated models and far more information to work with, households had a superior track record at forecasting inflation (10).

While the forecasting record of economists is bad, they are at least advantaged in that they only have to predict changes in macroeconomic variables that are relatively stable. In predicting changes in more volatile variables such as company and industry earnings, the forecasting record of the "experts" is downright atrocious.

In predicting the future earnings of companies, "professional" analysts are systematically inaccurate. In a business where a 5 or 6 percent "miss" of a projection can result in dramatic swings in the price of an associated company's stock, analysts' earnings estimates miss by over 15 percent on average (11). As can be seen below, in several studies carried out during the mid-1970's, the average error of analysts' earnings forecasts was 16.6%:

Notably, this high degree of error is not due simply to major errors on the part of a few analysts' forecasts. In most of the studies the variation among forecasts on each company was insignificant (12).

A study of longer-term earnings projections was carried out by Cragg and Malkiel and published in the Journal of Finance in March 1968 (13). The two professors examined the earnings forecasts of large groups of security analysts working for large investment organizations. Estimates were made for 185 companies for periods of from one to five years. The researchers found that most analysts' estimates were equivalent to linear extrapolations of current trends, and that correlations between actual and predicted earnings were very low. Thus, in spite of the vast additional information analysts gather, supplemented by frequent company visits, estimates were no better than forecasting a continuation of past trends. Remarkably, the reseachers found that more accurate forecasts could have been achieved by simply projecting a sample company's earnings to grow at the long-term average annual rate of 4 percent.

With regard to future earnings patterns in industries, expert forecasts are just as inaccurate as for individual companies. As can be seen below, in forecasting earnings for major industries between 1972 and 1976, analysts estimates were off by some 26 percent on average:

The reason that economists and business analysts are consistently inaccurate is because they apply selective and biased judgement in rendering conclusions and making projections. Unfortunately, the human psyche has very limited information-processing capabilities. In seeking to accurately assess current and future conditions in the highly complex, interdependent, and seemingly uncertain world of a modern market economy, economists and business analysts must integrate untold amounts of quantitative and qualitative information from a diverse set of sources. A virtually unlimited set of interacting factors shape the current state and future course of the economy and business earnings, and, what is worse, these factors are constantly changing or shifting in importance. Thus, regardless of how much information the so-called experts may be armed with, they are still defenseless against radical uncertainty. When it comes right down to it, economists and business analysts rely upon naive models and juqgemental heuristics that are unproven, unreliable, and subjective.

David Dreman sums up the situation this way:

-Market Irrationality and Value Investing-

"Markets have always had a manic-depressive quality to them, and whatever the direction in which
they are currently moving, investors expect 'sophisticated professionals' to lead the charge. "
(David Dreman, Psvchology and the Stock Market)

The subjectivity and inaccuracy of economists and business analysts is only one aspect of the irrationality found in markets. With regard to equities, investors systematically shift into poor performing assets and away from superior performing assets. More specifically, the investment crowd, influenced by "expert" advice that is consistently wrong, move into stocks that underperform the market and away from stocks that outperform the market.

David Dreman compiled 52 surveys from 1929 to 1980 that compared the performance of the experts' favorite company and industry stocks to the performance of the overall market. As can be seen below, if you had followed the experts' investment choices over that period, you would have underperformed the market 77 percent of the time:

That investors are drawn into poor performing stocks and away from the best performing stocks becomes most apparent by examining the effectiveness of "value" investing. Value investing is a strategy that involves purchasing stocks that have a low price relative to earnings. The basic measuring rod used to determine such value is the price-to-earnings ratio (P/E ratio): a stock's price per share divided by associated earnings per share. Historically, stocks with low P/E's relative to the rest of the market will outperform the market, and stocks with high P/E's relative to the rest of the market will underperform market averages.

In the following table, the annual performance of the 10 lowest P/E stocks in the Dow Jones Industrial Average is compared with the performance of the 10 highest P/E stocks in the DJIA as well as the performance of the DJIA as a whole:

As can be seen, low P/E stocks consistently outperformed the DJIA and and high P/E stocks significantly underperformed the DJIA.

What is most remarkable about value investing is that higher returns for buying low P/E stocks is possible with less, not more, systematic risk. As can be seen below, high P/E multiples entail higher investment risk and lower annualized returns whereas low P/E stocks offer lower risks and higher returns:

The reason that value investing works is because of market irrationality. Due to speculative greed and fear and the herd mentality of investors, popular stocks are bid up to irrational extremes while unpopular stocks are bid down to irrational extremes. While the stocks of given companies or industries are in play or out of play, so-called experts reinforce the crowd's irrational biases by offering advice the crowd wants to hear. In other words, the experts, hoping to win the hearts and, more importantly, the business of investors, consistently form rationales for, and make recommendations to, invest wherever the crowd is interested in investing anyways. Thus, individual stocks and industry groups repeatedly go through huge price gyrations around intrinsic value. By buying a stock that has a low price relative to an underlying measure of value such as earnings, an investor is going against the irrational bias of the crowd. Such a stock is unpopular, it is out of favor, and is therefore priced below intrinsic value and is near a cyclical low. The price eventually begins to come back, and, when it does, speculative greed will bring in a herd of investors that will send the market price beyond intrinsic value on the upside; thus, the truly "rational" investor will have a signal to sellout.

To specifically show how investors push market prices above and below intrinsic values, consider the case for the "investor-overreaction hypothesis". The investor-overreaction hypothesis holds that investors over- or under-value a stock after an earnings dissappointment. Because of this valuation error, truly rational investors are offered an opportunity to buy stocks at a discount. Evidence that investors overreact to earnings disappointments is found in a study carried out by Marine Midland Bank on the S&P 425 Industrials between 1948 and 1967 (15). As can be seen below, if you had purchased stocks reporting a loss one year after the deficit occured you would, over the course of the next one to five years, reap more than double the overall market return:

-Market Irrationality and Successful Market Timing-

While value investing is effective in determining which stocks to invest in at any given time, it is also effective at determining ~ to invest in stocks in the first place. More specifically, by paying attention to valuation measures of stock prices in general, an investor can become a successful market timer, i.e., an investor can consistently buy stocks near cyclical bottoms and sellout near market tops.

Consider, for instance, the graph below in which the historical swings in the average P/E ratio on the S&P 500 is compared with the S&P's price performance:

Figure 8

As can be seen, if an investor bought stocks whenever the market became relatively undervalued, i.e., the P/E on the S&P slipped below 10, then they would have bought near a major market bottom each time. Likewise, if an investor sold stocks every time the market became relatively overvalued, i.e., the P/E on the S&P climbed above 18, then, more often than not, they would have sold out near a major market top- the only exceptions being when the P/E ratio was inflated by depressed business earnings from major economic contractions such as during the 1930's. Notably, each time the market P/E has risen above 20 for two or more consecutive quarters without depressed earnings, a stock market crash has followed: this occurred in 1929, 1946, 1962, and 1987 (16).

Because P/E ratios are vulnerable to variations caused by temporarily depressed business earnings, a more consistent measure of intrinsic value that is used for successful market timing is the dividend yield.

Figure 9

The dividend yield of a stock is the annual dividend payed per share divided by price per share. If a dividend yield is relatively high, a stock is likely undervalued; if a yield is relatively low, it indicates overvaluation. Because dividends are less volatile than earnings, the average dividend yield of a market index is more stable than its average P/E ratio. As can be seen in the table above, by purchasing stocks when the average dividend yield on the DJIA is relatively high and selling when the average yield is relatively low, an investor can consistently beat the market.

While using valuation measures is one strategy for timing market tops and bottoms, another strategy is to systematically bet against the expectations of investors whenever a consensus opinion develops with regard to the future direction of market prices. This is the epitome of a contrarian investment strategy and it is how "contrarianism" earned its name. In order to determine the popular sentiment of investors, there are, fortunately, regular opinion polls of small investors, floor traders, institutional investors, and advisory services (17). These polls typically check whether someone is bullish or bearish; a bullish investor expects stock prices to head higher and a bearish investor expects market prices to head lower. Historically, whenever there is a consensus opinion among investors or investment advisors with regard to the future direction of market prices, i.e., whenever there is a majority bullish or bearish opinion, the stock market subsequently heads in the opposite, unexpected direction. In fact, the greater the consensus, the more substantially the market heads in the opposite direction. Below is a chart and table of the "forecasting" record of stock market advisors that show just how consistently wrong consensus market opinions are:

Figure 10

Figure 11

Importantly, polling data of individual investors, floor traders, and institutional investors shows the exact same pattern as above.

Figure 12

To get an idea of just how negative the correlation is between advisory expectations and subsequent market action, examine the following scatter plot diagram based upon the data above:

Figure 13

As can be seen, the greater the consensus becomes with regard to the future direction of the market, the more substantial is the subsequent price move in the opposite, unexpected direction. If investment advisors' expectations were truly "rational", as the efficient market hypothesis holds, then the line in the diagram above would be upward sloping or flat, reflecting that what investors expect consistently' proves true or insignificant. That the line is downward sloping establishes the irrationality of investment advisors (and the irrationality of the Efficient Market Hypothesis and economists who believe in it).

That irrational expectations lead to systematic investing errors is visible in the investment track record of mutual funds.

Figure 14

Figure 15

In seeking to maximize returns, mutual fund managers should seek to buy near stock market bottoms and sell near market tops. At the least, they should seek to avoid the market's intermediate-term ups and downs by buying and holding. Amazingly, however, fund managers do neither of these. Historically, mutual funds become almost fully invested near market tops and relatively underinvested near price bottoms. They are so good at this that one can systematically beat the market by investing in the opposite way that fund managers do. As can be seen above, when! mutual funds have a relatively low cash level, it indicates over optimism and a market top. When their cash levels are relatively high, it indicates over pessimism and a market bottom.

As can be seen in the scatter plot diagram below, there is a significant inverse relationship between mutual fund investment timing and subsequent market performance:

Figure 16

As you can see, the more fully invested mutual funds become, the more substantial is the subsequent stock market decline; the more underinvested mutual funds become, the more the market will subsequently rise. Mutual funds, representative of investors in general, make systematic errors when investing.

As a final example of contrarian indicators of market sentiment, consider the pace of secondary stock offerings.

Figure 17

As can be seen above, a significant inverse relationship exists between the rate of secondary stock offerings and the subsequent direction of the stock market. As speculation and optimism become excessive, a relatively large number companies will issue new stock. This is because such corporations are either caught up in the high expectations of the investment crowd and/or because they want to take advantage of low capital-raising costs associated with low-dividend yields (18).

All in all, an accurate description of investor and market behavior is achieved by bringing together the evidence above of swings in market valuation, investor expectations, stock fund cash levels, and secondary offerings. The reality is that investors, assumed to be objective and rational by efficient market theorists, are, in fact, subjective and irrational. Because each investor is drawn in and out of the market by selfish greed and fear, respectively, an insane crowd psychology is created that swings between extremes of greedy optimism and fearf\'.l pessimism. As market prices move upward, the greed mentality of investors leads to the formation of ever higher and unrealistic expectations. As expectations rise and an irrational group consensus forms, more and more money is invested and cash levels dwindle. Also, high expectations and relatively cheap capital leads more and more companies to seek capital through secondary offerings. As more money and investors are attracted to the market, turnover increases and prices are pushed to relatively overvalued levels. Eventually, the unrealistic optimism of the crowd reaches an extreme, a maximum of investors have been attracted to the market, and as much money as possible is committed; there is no place left for market prices to go but down. Once prices reverse, i~vestors respond with equal or even worse insanity on the downside. Once selling prevails and prices keep dropping, mass fear leads to the formation of increasingly unrealistic, pessimistic expectations. More and more investors sell their stocks and leave the market, cash levels rise, turnover decreases (following any initial "panic"), and prices are pushed to relatively undervalued levels. Eventually another extreme of unrealistic expectations is reached among investors, but this time of a negative character; a market bottom is eventually made and the stage is set for a new upswing in market prices.

-Market Irrationality as a Symptom of Social Insanity-

What is remarkable about the evidence above is that it shows only a small cross-section of a far more pervasive irrationality. Irrational mood swings are not limited to market psychology; they pervade the popular psychology of society as whole. To understand why, one only has to consider how contrarianism is applicable to sentiment indicators that go beyond gauges of investor opinion.

One example of the greater irrationality is the contrarian value of advertising and articles in popular magazines and newspapers. Martin Zweig, a major proponent of contrarianism, has noted that the general sentiment of advertisements in major periodicals is inversely related to the future course of markets and the economy. To time the stock market, he uses the number of bullish versus bearish ads in Barron's financial weekly as a contrarian indicator. When there is a disproportionate share of optimistic advertisments, the stock market subsequently heads lower and, when most ads are bearish, stock prices tend to head higher:

Figure 18

Another example of how popular magazines have contrarian value is the "Time-cover" indicator (19). Historically, whenever a major periodical such as Time magazine has a feature article on the direction of the economy and/or markets,. the opposite of what is expected in the article subsequently proves to be the case. Paul Montgomery, an analyst with Legg Mason Wood Walker Inc., researched this contrarian phenomenon using Time magazine cover stories going back to 1914. He found that when a financial topic is shown on the cover of Time, the trend or condition featured usually continued in the same direction for a month, but the situation in 11 months was the opposite of that indicated in the cover story 80 percent of the time. For example, in early November of 1987, Time ran a cover story entitled "The Crash". Consequently, Montgomery wrote the following letter to his clients:

When a Bear has been the cover story, as is the case today, the average market has continued to decline at a 30%+ annualized rate for at least a month before firming. Twelve months after a financial subject has made the cover of Time, the situation has been the opposite to the one suggested in nearly 80% of the cases -- outperforming a buy-and-hold strategy fivefold. If historic Time cover probabilities accurately represent current probabilities, the odds are nearly 2 to 1 in favor of new lows in stock prices in the corning weeks. But they are 3 or 4 to 1 that sometime between December and February, a dramatic sustained rally should begin (20).

This projection subsequently proved correct. Another example of the Time cover effect occured during the summer of 1988. On July 4th, 1988, a picture of a parched and barren grain field appeared on the cover of Time and the feature cover story, "The Big Dry", suggested that the Midwest drought of that summer would cause much higher bean prices. Just as the issue carne out, a sharp rally in bean prices that had been underway up until that time abruptly ended. With regard to interest rates, Montgomery considers using the Time cover indicator "the single most reliable method for timing long-term purchases and sales of the Bond marketll. During the 1980's, on each occasion that Paul Volcker appeared on the cover of Time, interest rates subsequently moved opposite the sentiment of the story.

The reason that magazine covers, articles, and advertisements have contrarian value is because social mood, not just the mood of investors, swings between optimism and pessimism. Because of this, all elements of popular culture reflect bipolar cycles in mood and have predictive, contrarian value. In 1985, Robert Prechter described the relationship between price swings in the stock market and swings in social mood in a special report entitled, 'Popular Culture and the Stock Market'. In this report, Prechter emphasized two points:

1. Popular art, fashion and mores are a reflection of the dominant public mood.
2. Because the stock market changes direction in step with these expressions of mood, it is another coincident register of the dominant public mood and changes in it.

Elaborating on the above points, Prechter wrote:

Both a study of the stock market and a study in popular attitudes support the conclusion that the movement of aggregate stock prices is a direct recording of mood and mood change within the investment community, and by extension, within the society at large. It is clear that extremes in popular cultural trends coincide with extremes in stock prices, since they peak and trough coincidentally in their reflection of popular mood... .The stock market is literally a drawing of how the scales of mass mood are tipping. A decline indicates an increasing "negativel' mood on balance, and an advance indicates an increasing "positive" mood on balance. (23)

According to Prechter, popular mood swings are registered in basically all cultural phenomena: "Trends in music, movies, fashion, literature, television, popular philosophy, sports, dance, automobile styling, mores, sexual identity, family life, campus activities, politics and poetry all reflect the prevailing mood, somtimes in subtle ways. (24)" Thus, cycles in social mood can be gauged not only by monitoring trends in stock prices and investor sentiment but also by keeping an eye on trends in popular culture. In fashion, a rising social mood is reflected in rising hemlines and more vivid colors, like during the 1960's; a declining mood is reflected by lengthening hemlines and more drab colors, such as occurred during the 1970's. In movies, popular optimism entails a prevalence of upbeat films like the musicals of the 1960's, whereas mass pessimism is associated with horror films such as during the 1930's when movies were about King Kong, Frankenstein, vampires, werewolves, and mummies. In popular music, the positive or negative trend in social mood is strongly apparent. This can be seen by examining figure 19 (on the next page) in which cyclical trends in stock prices are charted along side dominant trends in pop music.

All in all, swings in stock prices and market irrationality is but a symptom of irrational swings in the prevailing mood of society as a whole. Contrarianism is applicable to cultural trends in the same way that such an approach is used to anticipate future price trends on Wall street. Since social mood repeatedly swings between optimism and pessimism, and because the prevailing mood shapes popular culture, the general sentiment of fashion, movies, music, and other such cultural phenomenon is somewhat predictable. In other words, social change and cultural trends are shaped by collective emotional and, in turn, mental instability. Thus, just as market irrationality allows a contrarian to anticipate future movements in stock prices, social insanity allows a contrarian to anticipate future trends in popular culture.

Figure 19