Trade Like an O'Neil Disciple: How We Made 18,000% in the Stock Market

7.1. TIMING THE MARKET

A timing system must start by containing internal logic that makes sense; then the system can be built around this internal logic. This is where many years of market experience are necessary. This avoids the black box situation of overfitting data where data is fit to predict the past but has little to no predictive value on the future. Unfortunately, so many timing systems available on the Internet for a subscription fee lack internal logic but manage to boast high theoretical returns because they have overfitted their past data. Such systems are great at predicting the past but fail when tested under real-time market conditions. This is perhaps why such skepticism abounds when the words "timing model" or "timing the market" are mentioned. But just because most timing systems do not work, does not mean all timing systems do not work. We highly recommend Robert Koppel's book Bulls, Bears, and Millionaires (Dearborn Financial Publishing, 1997), where the author and Mike Dever discuss the perils of model design and overfitting data.

I have been timing the markets real-time, under fire since 1991, the first successful year where I strongly outperformed the major U.S. market averages. My model is a statistical formalization of price/volume action of the NASDAQ Composite and S&P 500. I developed a set of rules that guides this model and has guided my trading over the years. It is largely responsible for my long-standing track record, as verified by the big four ...