Variable Cycle Lengths And Intraday Candlestick Analysis by Gary S. Wagner and Bradley L. Matheny
Trading methods often use different length filters for generating trades, both entries and exits. The most
common filters are moving averages with different lookback periods. Here, STOCKS & COMMODITIES
contributors Gary Wagner and Brad Matheny apply different filter lengths to candlestick charts by using
different period lengths of candlestick charts for confirmation of bullish and bearish patterns.
In "Candlesticks and intraday market analysis," we looked at the signals generated using a 22-minute
candlestick chart of the March 1993 Standard & Poor's 500. That article illustrated the candlestick
patterns generated with intraday data. Although simplistic in design, the 22-minute filter provides valid
results. However, there are two drawbacks in this technique. First, a 22-minute cycle will not be sensitive
enough to capture the very short-term price swings that occur during the day. Second, the entry and exit
points may not be as timely as the ones obtained from the use of multiple time cycles.
The 22-minute cycle study is the first step of a more intricate technique that attempts to capitalize on
short-term price fluctuations. This technique compares candlestick patterns produced by multiple time
cycles and only takes signals if two or more cycle lengths confirm each other. If correctly used, this
approach can increase your probability of success.
Trading signals generated from the shorter cycle length are more sensitive to imminent market changes,
but they are also more likely to generate false calls. Taking only the signals that are confirmed by longer
cycle lengths will filter out many of the false calls produced by a shorter time cycle. In this way, the edge
gained with the use of very short cycles will be balanced with the reliability of longer time cycles, thereby
producing more effective results.