We hear it on CNBC regularly… a certain stock has shown a “golden cross” pattern. This could be the beginning of a big bull trend. Another stock has shown a “death cross” pattern, and now is definitely the time to get out. These vaguely-defined strategies are how many people perceive technical analysis.
Often, the strategy is applied in the same way, with the same variables, to a variety of different markets. The individuals using it expect to learn something of value. Is the fast MA 50, 25, or 100? Is the slow MA 100, 150, or 200? Do you sell buy as soon as the cross occurs, or do you wait for a pullback? Everyone has their own interpretation, and anyone who knows the Chimps knows we don’t leave room for interpretation.
I propose an experiment. Let’s give the Golden Cross and the Death Cross free rein to define themselves on their own terms. We’ll give them full leniency, and let them perform as well as they possibly can. No stop loss, and history from 1949 on the S&P 500. We’ll let the fast and slow moving averages optimize themselves. Go go gadget Genesis…
And the results are in! I’ve highlighted a period between 2009 and 2012 during which a series of crossovers occurred. Barely visible are the buy and sell triggers, for example, you can see a small red triangle between June and July 2010 right after a crossover during which the system closed its long position and opened a short.
The optimal values for fast and slow MAs on the S&P 500 from 1949 to present are 56 fast and 198 slow. Actually, not too far off from one I’ve often heard, 50 fast and 200 slow. But before you go programming this into your own charting program, let’s take a closer look at performance…
Thinking twice about applying this strategy to your own trading? You’d better be. You’ll find yourself making money on just 1 out of 5 trades. With such a poor track record, you’d be a trader of amazing resolve if you were able to make that 4th or 5th trade believing it’s likely to be a winner. The numbers speak for themselves. This is clearly not a strategy anyone should be considering.
But the question remains, is there any way to salvage a wreck like this? The Chimps think so. Even the most standard and rudimentary (dare we say, fundamental?) of technical methods like Moving Averages must be applied correctly. Next week, we’ll show you how to we modify Moving Averages to come up with insanely better results.