by John F. Ehlers
The first thing we must recognize is that cycle analysis doesn't always work for profitable investment
strategy. This is no great surprise, because cycles must be present in the data history for any analysis of
them to be valid. Also, to be useful, we must assume that the cycles in the history will extend into the
future so that a prediction can be made. This is sometimes impossible because of fundamental issues. For
example, a freeze in Florida will probably swamp any technical factors on orange juice futures. Even
within these constraints for technical analysis, cycles can still be difficult.
This is the first of a series of articles on cycle analysis. In this article, we will go back to basics and
briefly review cycle theory to have a common basis for definitions. The program listing in the panel gives
an ideal trading system for a perfectly cyclically varying price. More importantly, the program is
interactive to show the impact of dominant cycle resolution and spectral purity on the accuracy of the
trading system. In future articles, we will show how some of the traditional trading indicators can be
optimized through the use of cycle analysis.
If you stop and think about it, you might ask "Why did Welles Wilder use 14 days in his RSI? Is there a
better period to use?" It turns out that the best period to use is related to the dominant cycle of the data
history. Future articles will have discussions and program listings for the Relative Strength Indicator
(RSI), Daily Trend Indicator (DTI) and Commodity Channel Indicator (CCI).
After studying this article you should be able to recognize when cycles are present, how to estimate the
dominant cycle without extensive analysis, and whether there is sufficient spectral purity (1) to make
cycle analysis useful.