Trading indicators are part of every technical traders arsenal. Yet I’ve found that most traders and investors are discouraged with most indicator’s performance. I think the reason is because they are trying to use the indicator like everyone else. Stop and think, if the conventional application was profitable why do so many traders lose money?
Most profitable traders I know have found an edge and exploit it to their advantage. You would probably be astonished to discover the simplicity of many trading “edges”.
This two part article will reveal the plain and honest truth about a popular indicator, the MACD. Today, I want to describe the MACD and show how it is most ofter used. In Part II, I will offer a different application for your consideration.
You’ve probably heard that Gerald Appel invented the MACD or moving average convergence divergence. Commonly called the “Mack -D”, it is an exponential moving average (EMA) that uses the difference in values between two EMAs.
The default or standard MACD is a 12 day EMA minus a 26 EMA. A 9 period EMA of the MACD is plotted as a signal line. The histogram (the white bars on the following chart) is the difference between the MACD and the signal line. It measures the degree of momentum.
The common way to use the MACD is to make buy and sell decisions based on the fast line crossing above (bullish) or below(bearish) the signal line. At first glance these signals look very interesting.
This chart is the same chart as above — the data is a little more compressed showing March 2009 to the current July 2010. I’ve run a simple script asking for buys and sells of the MACD/signal line crosses. The green arrows are buys, the red arrows are sells. Here are the results:
This is the back test results covering the period visible on the chart — April 2009 through June 2010. The actual trade dates, price, etc. are on the grid on the right.
I ran the same test on a list of stocks considered to be bellwether stocks. For the same time period, these are the results:
I don’t need to run this back test over more time frames to know that this strategy is not up to my minimum standards.
In Part II I’ll suggest a better way to use the MACD.
