Van Tharp on developing a trading system

http://www.iitm.com/Weekly_update/Weekly_289_Sept_20_2006.htm


Randomness.
People want to treat the world as if they could predict and understand everything. As a result, one of the most significant biases people have is to seek patterns where none exist and to invent the existence of unjustified causal relationships. Traders don’t want to trade probabilities. They want consistency. For example, people fail to understand that a random sequence can include a long string or what would be called a trend. Instead, they try to understand the “trend” as something that it isn’t, instead of accepting that such phenomena occur.

Understanding and trading well are not necessarily the same thing. People don’t understand randomness, yet they expect to be able to understand the market. They then build trading systems out of their attempts to understand the market by identifying unjustified causal relationships without ever realizing they are doing it. It is this expectation to understand markets that leads traders to search for “Holy Grail” trading systems that explain the “underlying order” of the markets. There is nothing wrong with building a trading system based on microcosmic glimpses into how the market might work; but you need to know what you’re doing when you’re doing it. You are not trying to understand some mysterious underlying order in the markets. You are developing a set of rules whose long term expectancy gives you an edge in the market, while allowing you to withstand the worst possible catastrophe that could occur in the short term.

For example, many people observe a relationship in the market and assume it explains how the market works.

Jack noticed when a particular pattern occurred in the market, it frequently moved 50 to 100 points higher. He assumed the pattern meant that strong hands were moving into the market. And, when the market didn’t follow the pattern, he became very confused. I said, “How often, when you observe this pattern, does the market move like that?” He responded, “About 35% of the time!” Thus, Jack had simply observed a pattern that was quite profitable 35% of the time. The rest of the time it had no meaning.

A relationship may occur only 35% of the time, and that may be something you can make money with, but it has nothing to do with being right or trying to explain something. What you must learn is that most trading systems come out of observations that have a certain probability of being correct. Those observations do not explain anything. Remember, a trading system is just a set of rules to guide behavior, nothing less or nothing more. Apparent random fluctuations in the market are caused by many more factors than you can possibly monitor in your system.

Develop the attitude of following rules
because they give you an edge in the market.
Avoid the need to understand or explain the market.

Because people attempt to understand and make order out of the market, they assume that the longer a trend continues, the more likely it will suddenly turn around. More importantly, traders are usually willing to bet larger amounts of money on that assumption. Thus, traders want to pick tops and bottoms in a trend—a behavior that tends to be as dangerous as stepping in front of a moving freight train, hoping it will stop and turn around just for you. These biases are usually referred to as the gambler’s fallacy. They have resulted in the ruin of millions of traders over the ages. The gambler’s fallacy is one of those biases, which make trading difficult without a system and proper money management. However, traders frequently develop counter-trend following systems because of this bias—usually with disastrous results.