Friday, January 21, 2011

Pattern Recognition Criteria






While looking at a chart can often tell you whether a stock is appropriate for swing
trading, it is very time consuming to look at charts, particularly if you look for
opportunities every day.  Another way to identify good stocks is to use software that
can scan all of the listed stocks based on a series of algebraic equations that
represent the characteristics of a good chart pattern.  I use SwingTracker to
accomplish this task.

Before discussing the specifics of pattern recognition criteria, we’ll briefly
consider the measures used in the algebraic equations.  Some of the measures are
simple descriptive variables (e.g., the high price for the previous day  or the average
volume over the past 20 days).  Other measures are based on technical analysis
which is discussed in more detail in the Appendix.  Technical analysis has many
different indicators from a simple moving average to a complex oscillator.  It is not
necessary to have an in-depth understanding of technical analysis to be a successful
swing trader, however, it is helpful to have a rudimentary understanding of how we
approach swing trading pattern recognition.



Technical Analysis Measures used to Recognize Swing Trading Patterns 


To begin with, we typically restrict our selections to stocks that are at least $12 in
price, having an average (20 day) daily volume of at least 500,000 shares.  Since
market makers can more easily manipulate low price, low volume stocks, we stay
away from them.
For long swings we are interested in identifying stocks that are in an uptrend.  One
of the indicators we use is a simple moving average (SMA).  A moving average is
simply the average closing price for a particular number of days.  It’s called a moving
average because on each new day, the current day’s price is added to the average
while the oldest price is dropped.  We typically focus on three moving averages,
those based on 10 days, 20 days and 50 days.  All moving averages smooth the
price movement and make it easier to identify trends.  It is also significant to know
where today’s price is relative to the moving averages and whether the shorter timeframe moving average is above or below the longer time-frame moving average.
Two indicators that a stock is in an uptrend are:


•  Today’s closing price is above both the 10-day and 20-day moving averages
•  The 10-day moving average is above the 20-day moving average
When looking for a long swing, we would like to identify stocks that are
experiencing a brief decline (pullback).  We can identify a 3-day pullback as follows.
•  Today’s high price is lower than yesterday’s high
•  Yesterday’s high is lower than the high the day before
We also use a technical indicator developed by Dr. Alexander Elder called the Force
Index.  This index combines the magnitude of the price change with the direction of
the change and the trading volume.  In order to confirm the relative force behind an
uptrend and a pullback, we use a 3-day moving average and a 13-day moving
average of the Force Index.  The following conditions demonstrate that the bears
have been winning the short-term battle while bulls are dominating the longer
frame:
•  The 3-day moving average of the Force Index is less than 0, and
•  The 13-day moving average of the Force Index is greater than 0
Another technical indicator we like to use is the Directional Movement Index (DMI)
that was developed by J. Welles Wilder Jr.  It is used to determine whether a stock is
trending or not trending (i.e., moving sideways).  In SwingTracker we provide the
two components of this indicator – the Positive Directional Index (+DI) and the
Negative Directional Index (-DI) – along with a 20-day moving average based on
these two measures (ADX).  An uptrend is confirmed if …
•  ADX is higher than 30
•  +DI is greater than –DI