CHAPTER 2

Is It All Just Random?

According to many academics, technical analysis is a pure waste of time. Price, they claim, is absolutely random. Using patterns and indicators to predict its behavior is no different and no less primitive than reading entrails of a dead animal in order to divine your future. One of the favorite tricks of the pure randomness crowd is to have the random function in Microsoft Excel generate a series of numbers and then plot it on a graph. Admittedly, when many technically oriented traders are confronted with a seemingly nice chart pattern only to be told later that it's all random, they experience a loss of confidence. Is it all a ruse? Is technical analysis useless? Are we hopelessly wasting our time trying to learn its precepts? No, no, and no.

Price patterns are no more random than all human behavior—that is to say, they can be quite accurately predicted in general but quite often miss in the specific. Let's play the following game. Your job is to observe a wealthy Upper East Side businessman (for fun we'll assume that he always wears a dark blue pin-striped suit and yellow polka-dotted tie and is known to all as Mr. X) as he leaves his Park Avenue penthouse to walk 20 blocks to his high-rise office in the famous Met Life Insurance building. If you can accurately predict when he will appear in the door of his apartment building you win $1,000. If you are wrong you lose $1,000. What do you think your chances of winning would be if you simply had to choose morning or afternoon? How about if you were required to call the time to the hour? What about to the minute? The second? The millisecond? Of course, the more precise you had to be the less likely you would be right and the more likely you would lose money. File that thought away, because we will come back to this very important concept later in the book. In the meantime, answer this question: Are the activities of Mr. X random just because you cannot consistently predict his appearance at the door of his apartment building to the second or the millisecond of occurring? Of course not.

In fact Mr. X's behavior is highly predictable. If you observe him long enough you might know that on Fridays he seldom ventures out, preferring to work at home. Or that during Christmas week he doesn't go to the office at all, leaving the city altogether. Or perhaps that on particularly pleasant mornings he likes to linger on his balcony and smell the flowers planted in his flower boxes before heading off to the office. You would obtain all this information through observation—which is what speculation really is. The more you observed the better your information would become. Yet would you ever be able to bet your life savings on any one outcome of Mr. X's behavior? No, not if you were smart—because all of Mr. X's actions, like all human behavior and like all life, has an element of chaos to them. One Christmas week, for example, he may have been involved in a multibillion-dollar deal and therefore gone to his office every day. Another time he may have been bedridden with the flu and not have come out for days. Yet another day a neighbor's dog might have unexpectedly jumped on him in the elevator and torn his pants, necessitating a trip back upstairs and delaying his appearance at the door by more than 15 minutes.

To an untrained novice unaccustomed to Mr. X's habits, these changes of behavior would connote randomness and the player will most likely give up on the game after losing several thousand dollars, arguing that Mr. X's actions are completely unpredictable. But a skilled observer would understand that these deviations are simply the result of the normal degree of chaos in Mr. X's life. An expert in Mr. X's movements would know with a high degree of certainty that on most mornings Mr. X would appear at the door of his apartment building before 8 A.M., and the expert would be able to collect $1,000 making that bet.

Price data exhibits very similar dynamics. Unlike ivory-towered academicians, veteran traders who actually observe price action day in and day out for thousands of days at a time, realize that what is most striking about price behavior is not its wild randomness, but rather its mundane repetitiveness. Price patterns repeat themselves over and over and over again on many different time scales. What is different each time—and what makes trading sometimes maddeningly frustrating—is the amplitude of the move. Sometimes breakouts can last for 100 points and sometimes only for 10—just as sometimes Mr. X will have 100 uninterrupted days of punctually walking to his office and sometimes his schedule will be wildly skewed as other events in his life cause a temporary change of pattern.

Observation through technical analysis simply provides the trader a degree of expectancy but absolutely no guarantee that any particular trade analysis will be correct. Nevertheless, the greater the degree of expectancy, the better the trader's edge and the stronger the chance of ultimate success. After all, markets are not some robotic mechanisms that can be studied with keen dispassion through some elegant mathematical models. They are living, breathing organisms made up of millions of traders.

What do we trade in FX? I often ask this question at seminars. I get many answers but rarely the right one, for what we trade in FX is what is traded in all markets—sentiment. Fundamental factors shape and manufacture sentiment, while technical analysis expresses it. Price patterns are simply the reflection of repetitive human reactions of fear and greed to ever-changing news flow. They are not some randomly generated numbers from an Excel spreadsheet, even though they may look very similar. This is the critical flaw of pure randomness theorists—just because random functions can often mimic price patterns does not mean that price patterns themselves are random.