Having the perfect indicator would be one thing, but putting it into practice would be another. Even if you are able to "beat the market" the greater battle of "beating yourself," that is, mastering your emotions, still lies ahead. Every great market operator, whether a trader or an investor, knows that the analytical aspect of playing the market represents only a small segment compared with its psychological aspect. In this respect, history's great traders or investors-to one degree or another-have followed various rules. However, these successful individuals would be the first to admit that they have no convenient magic formula to pass on as a testament to their triumphs.
The false "Holy Grail" concept appears in many forms; we will consider two: the expert and the fail-safe system, or perfect indicator.
The Myth of the Expert
All of us gain some degree of comfort from knowing that we are getting expert advice whenever we undertake a new task. This is because we feel somewhat insecure and need the reassurances that an expert-with his undoubted talents and years of experience-can provide. However, it is not generally recognized that experts, despite their training and knowledge, can be as wrong as the rest of us.
It is always necessary to analyze the motives of experts. Britain's Prime Minister Neville Chamberlain, having returned from Hitler's Germany with a piece of paper promising "peace in our time," no doubt believed wholeheartedly the truth of his grand statement. The fact was, he was an expert, and he got it wrong. President John Kennedy also had his problems with experts. "How could I have been so far off base? All my life I've known better than to depend on the experts," he said shortly after the Bay of Pigs fiasco.
Classic errors abound in military, philosophical, and scientific areas. In the investment field, the record is perhaps even more dismal. One of the differences that sets aside market forecasters from other experts is that market prices are a totally accurate and impartial umpire. If you, as a financial expert, say that the Dow-Jones average will reach 3,500 by the end of the month and it goes to 2,500, there can be little argument that you were wrong. In other fields, there is always the possibility of hedging your bets or making a prognostication that can't be questioned until new evidence comes along. Those experts who for centuries argued that the world was flat had a heyday until Columbus came along. It didn't matter to the earlier sages; their reputations remained intact until well after their deaths. However, conventional thinkers after 1493 did have a problem when faced with impeachable proof.
Experts in financial markets do not enjoy the luxury of such a long delay. Let's take a look at a few forecasts. Just before the 1929 stock market crash, Yale economist Irving Fischer, the leading proponent of the quantity theory of money, said, "Stocks are now at what looks like a permanently high plateau." We could argue that he was an economist and was therefore commenting on events outside his chosen field of expertise. In the previous year, however, he also reportedly said, "Mr. Hoover knows as few men do the terrible evils of inflation and deflation, and the need of avoiding both if business and agriculture are to be stabilized." Up to the end of 1929, both were avoided, yet the market still crashed.
When we turn to stock market experts, there is even less to cheer about. Jesse Livermore was an extremely successful stock operator. In late 1929, he said, "To my mind this situation should go no further," meaning, of course, that the market had hit bottom. Inaccurate calls were not limited to traders. U.S. industrialist John D. Rockefeller put his money where his mouth was: "In the past week (mid-October 1929) my son and I have been purchasing sound common stocks." Other famous industrialists of the day agreed with him. One month later, in November 1929, Henry Ford is quoted as saying, "Things are better today than they were yesterday."
Roger Babson, one of the most successful money managers of the time, had in 1929 correctly called for a 60 to 80 point dip in the Dow. Yet, even he failed to anticipate how serious the situation would become by 1930, for he opined early in that year, "I certainly am optimistic regarding this fall. . . . There may soon be a stampede of orders and congestion of freight in certain lines and sections." Unfortunately, the Depression lasted for several more years. Perhaps the most astonishing quote comes from Reed Smoot, the chairman of the Senate Finance Committee. Commenting on the Smoot-Hawley Tariff Act, generally believed to be one of the principal catalysts of the Great Depression, he said, "One of the most powerful influences working toward business recovery is the tariff act which Congress passed in 1930." Figure 1-1 depicts market action between 1929 and 1932, thereby putting these experts' opinions into perspective.
The testimony of these so-called experts shows that some of the greatest and most successful industrialists and stock operators are by no means immune from making erroneous statements and unprofitable decisions. Common sense would have told most people that the stock market was due for some major corrective action in 1929. It was overvalued by historical benchmarks, speculation was rampant, and the nation's debt structure was top-heavy by any standard. The problem was that most people were unable to relate emotionally to this stark reality. When stock prices are rising rapidly and everyone is making money, it is easy to be lulled into a sense of false security by such "expert" testimony.
Of course, some individual commentators, analysts, and money managers are correct most of the time. We could, for instance, put Livermore and Babson into such a class. However, if you find yourself blindly following the views of a particular individual as a proxy for the Holy Grail, you will inevitably find yourself in trouble-probably at the most inconvenient moment.
An alternative to using a single guide is to follow a number of different experts simultaneously. This solution is even worse because experts as a group are almost always wrong. Figure 1-2 compares Standard and Poor's (S&P) Composite Index with the percentage of those writers of market letters who are bullish. The data were collected by Investor's Intelligence* and have been adjusted to iron out week-to-week fluctuations. (A more up-todate version appears in Chapter 8.) Even a cursory glance at the chart demonstrates quite clearly that most advisors are bullish at major market peaks and bearish at troughs. If this exercise were conducted for other investments such as bonds, currencies, or commodities, the results would be similar. At first glance, it may appear that you could use these data from a contrary point of view, buying when the experts are bearish and selling when they are bullish. Unfortunately, even this approach fails to deliver the Holy Grail, because the data do not always reach an extreme at all market turning points. At a major peak in 1980, for example, the Index couldn't even rally above 60%. In late 1981, on the other hand, the Index did reach an extreme, but this was well before the final low in prices in the summer of 1982. While the Advisory Sentiment Indicator does forecast some major peaks and troughs, it is by no means perfect and certainly lacks the consistency needed to qualify as the Holy Grail.
Figure 1-1 U.S. Stock Market 1927-1932. Source: Pring Market Review.
The Myth of the Perfect Indicator
It is almost impossible to flip through the financial pages of any magazine or newspaper without coming across an advertisement promising instant wealth. This publicity typically features a computerized system or an investment advisor hotline that claims to have achieved spectacular results over the past few months or even years. Normally, such services specialize in the futures or options markets because these highly leveraged areas are in a more obvious position to offer instant financial gratification. The huge leverage available to traders in the futures markets significantly reduces the time horizons available to customers. Consequently, the number of transactions, (i.e., revenue for the brokers) is that much greater.
Figure 1-2 S&P Composite versus Advisory Service Sentiment 19741984. Source: Pring Market Review.
As a rule of thumb, the more money the advertisement promises, the more you should question its veracity. History tells us that it is not possible to accumulate a significant amount of money in a brief time unless you are extremely lucky. Moreover, if you are f
ortunate enough to fall into a situation where the markets act in perfect harmony with the system or approach that you have adopted, you are likely to attribute your success to hidden talents just discovered. Instead of walking away from the table, you will continue to be lulled back into the market, not realizing the true reason for your good fortune. You will inevitably fritter away your winnings trying to regain those lost profits.
Consider the advertisement's promises from another angle. If the system is so profitable, why are its proponents going to the trouble of taking you on as a client and servicing your needs? Surely, it would be less bothersome to execute a few orders each day than to go to the trouble, expense, and risk of advertising the service. The answer is either that the system doesn't work or, more likely, that it has been tested only for a specific period in the most recent past. You, as a prospective user, should focus on the likelihood of the method's operating profitably in the future and not on some hypothetical profits of recent history.
Most systems base their claims of success on back-tested data in which buy-and-sell signals are generated by specific price actions, for example, when the price moves above or below a specific moving average. It seems natural to assume that past successes can forecast future profits, but the results of back-tested data are not as trustworthy as they appear. First, the conditions in which the data are tested are not the same as a real market situation. For example, the system may call for the sale of two contracts of December gold because the price closed below $400. On the surface, this may seem reasonable, but in reality it may not have been possible to execute the order at that price. Quite often, discouraging news will break overnight causing the market to open much lower the next day. Consequently, the sale would have been executed well below the previous $400 close. Even during the course of the day, unexpected news can cause markets to fluctuate abnormally. Under such conditions, systems tested statistically under one-day price movements will not reflect a reasonable order execution. An example of this situation arises when market participants are waiting for the Commerce Department to release a specific economic indicator. Occasionally when the announcement falls wide of expectations, a market will react almost instantly, often rising or falling 1% or 2%. The time frame is so short that it is physically impossible for many transactions to take place. As a result, the system does not truly indicate a realistic order execution.
Another example is the violent reaction of the market to some unexpected news. On the evening of January 15, 1990 (Eastern Standard Time), U.S. and allied troops began the invasion of Kuwait. The next day the market, as measured by the Dow-Jones average, rose well over 75 points at the start of trading. In effect, there was no opportunity to get in (or out if you were short) anywhere near to the previous night's close. This is an exceptional example, but it is remarkable how many "exceptions" occur as soon as you try to adapt one of these methods to the actual marketplace.
Another flaw with these systems is that data are usually back tested for a specific time, and special rules are introduced so that the method fits the data retroactively solely to demonstrate huge paper profits. If you invent enough rules, it is relatively easy to show that a system has worked in the past. However, if rules are developed purely to justify profits in these specific periods, the chances are that these same rules will impede future success.
To ensure that a system is likely to work in the future, when it counts, the rules should be simple and kept to a minimum, and the testing period should cover many markets over many years. The problem with most of these advertised ventures is that they give you the results of only the most successful markets. If you ask the advocates of these schemes to report their findings for other time periods or other markets, you will be greeted with blank stares.
A final drawback of systems is that they usually fail when rolled out into the real world. The reason? Market conditions change. Figure 1-3 shows a system based on a simple moving average crossover. This method works well when the market shows a clear-cut trend of the kind seen between January and March 1991. However, the same system could hand you your head on a platter when price action is more volatile, as it was between midMarch and May 1991.
Figure 1-3 S&P versus a Twenty-Five-Day Moving Average. Source: Pring Market Review.
Changes in the character of a market are not just limited to changes in trend volatility. Any method that uses the past to forecast the future assumes that past behavior will repeat.
Systems constructed from assumptions concerning basic economic fundamentals are also subject to failure. For example, it has been established that, in almost all cases, stock prices sooner or later rally in the face of falling interest rates and begin to fall sometime after rates have begun to rise. The lags fall into a fairly predictable range most of the time but on occasion can be unduly long. These exceptions can result in missed opportunities or devastating losses. This problem occurred at the beginning of the Depression. Interest rates peaked in the fall of 1929, yet the stock market declined by about 75% over the next three years. In this instance, the knowledge that rates lead equity prices could have led to devastating losses. Timing is everything. In a similar vein, short-term interest rates bottomed out in December 1976 at 4.74% and almost quadrupled to a cyclical peak of 16.5% in March 1980. Yet stock prices in the same period as measured by the S&P Composite were unchanged.
While the inverse relationship of interest rates to equity prices works well as an indicator of market direction most of the time, these examples show that it is far from perfect and certainly no Holy Grail. The reason for this is that once a certain indicator or investment approach works for a while, word of its money-making capabilities spreads like wildfire. Then, when everyone is aware of its potential, it becomes factored into the price and the relationship breaks down.
This concept works just as well in reverse, where fear rather than greed is the motivator. People, it seems, tend to repeat past mistakes but not those of the most recent past. Once-bittentwice-shy applies as much to trading and investing as to any other form of human activity. In the 1973-1974 bear market, for example, equity investors were clobbered principally due to rising interest rates. In virtually every business cycle throughout history, investors have waited to sell stocks after interest rates started to rise. In the cycle that followed the 1973-1974 market debacle, however, investors sold stocks in anticipation of rising rates. Since rising interest rates were already factored into equity prices, the stock market actually rallied along with rates in the 1978-1979 period.
In New Methods for Profit in the Stock Market, Garfield Drew tells us that the mind generally harks back to its last experience in the market and judges the market by that encounter. He shows that stung investors hold postmortems after each unforeseen collapse in prices to get a better grip on the warning signals that preceded the collapse. People then concentrate on these factors to prepare themselves for the next failure. Drew points out that this is seldom a smart idea, because the dangers have shifted to another sector of the economy by the time the next decline begins.
An excellent example of this tendency occurred in the 1920s. In 1921, the U.S. economy suffered a short depression stemming from a rampant commodity inflation. As a result, most people felt that the situation in 1929 was sound because commodity prices were subdued. What they did not realize was that the problem of inflation had moved from commodities to broker loans, that is, margin accounts. When these unsound debts were involuntarily unwound due to weak market conditions, they acted as a catalyst for the downward spiral in stock prices in late 1929.
At the peak of the next cycle in 1937, investors were again complacent, because the amount of margin debt was low compared with 1929. The problem had moved once again. This time the cause of the subsequent recession and bear market in equities was inflated inventories in the face of a declining demand.
Drew goes on to describe the popularity of business barometers designed to identify turning points in the business cycle and the stock market. F
or the most part, these models had worked extremely well, but they were based on historical data. However, they failed completely in the late 1920s and early part of the 1930s, because that experience defied any historical norm. Almost without exception, these barometers indicated that stocks were cheap in 1930 and should be bought. This advice preceded the final low by about two years, when prices had yet to undergo a decline of 50%. Drew continues: "But 'normal' at any given time merely means an average of the past. It does not allow for changed conditions, whereas the current or future 'normal' may be something quite different."
Traditional relationships may break down in other ways because of institutional or technological changes. For many years, technical analysts used to track the number of shares sold short (a bearish maneuver made by speculators who believe they could profit from lower stock prices). A bullish reading occurs when the volume of the total short position on the New York Stock Exchange is greater than twice the average daily trading volume. This high ratio implies that the public is bearish, because it is shorting heavily. Also, because every share sold short has to be repurchased, a high ratio also indicates a huge potential demand for stock.
The short interest ratio worked beautifully, signaling nearly all the major bottoms between the 1930s and the early 1970s. At that time, stock options and stock index futures were introduced and much of the subsequent short selling represented hedge positions against these derivative products. These transactions have had the effect of distorting the short-interest data so that a high short-interest ratio no longer reflects pessimism or potential demand to the extent that it had done previously.
Investment Psychology Explained Page 2