Seminar

	STREET SMART SCHOOL The Classrooms
A service of StreetSmartReport.com

Classrooms Home Street Smart Report Home

Seminar #3.

TRADING BANDS.

In the last seminar we said moving averages are used to smooth out the market's up and down fluctuations within a trend without destroying the picture of whether an index (or stock) is in a rally or a correction. And we pointed out that moving averages usually provide support for pullbacks in a rising market, and overhead resistance for brief rallies in a declining market (until the trends reverse direction).

Trading bands placed around a moving average can provide additional useful information and guidance.

One that we particularly like is an 8% trading band around a 21 day moving average applied to the Dow.

The 8% band is produced by first calculating a 21 day simple moving average, the dotted red line on the above chart. A line that will be the upper limit of the trading band is produced by adding 4% to the value of the 21 day m.a. each day and plotting it on the chart. The lower limit of the trading band is then produced by subtracting 4% from the value of the 21 day m.a. each day and plotting it on the chart.

The above chart gives a long range picture of how the Dow rarely trades outside of the band, and when it does it almost always indicates an oversold or overbought condition and results in an immediate sharp reversal in the opposite direction. The same situation has prevailed for at least the last 30 or 40 years.

The following chart covers a shorter time span so we can see the activity in more detail.

First let's look at the activity in the rallies. Note that just as the 21 day moving average tends to provide support on each of the brief pullbacks within the rising trend, so the upper limit of the trading band usually provides an upside limit before the next brief pull-back to the m.a. begins. And so the Dow rises up the narrow channel between its 21 day m.a. and the upper limit of the trading band. The activity is even more accurate than presented in this chart since the chart is based on closing prices at the end of each day. With the intra-day volatility in the market in 1998 and 1999, with 100 point intra-day swings in the Dow common, even where on this chart the Dow seems to have not quite reached either the moving average or the trading band line before reversing, more often than not intra-day it did.

Similarly, when we look at the activity in the downturns, the 21 day m.a. tends to act as overhead resistance each time the market attempts to rally within a downtrend. And the lower limit of the trading band tends to act as temporary support in the decline. Once the lower limit of the band is touched a temporary bounce back up to the 21 day m.a. frequently takes place. And so the Dow declines down a channel between its 21 day m.a. and the lower limit of the trading band.

And as mentioned, on the rare occasions when the Dow breaks outside of the trading band it scrambles to get itself right back into the band.

It takes considerable computer time and experimentation to find the best mix of moving average and percentage trading band for different indexes, and for some there is no usable consistency. A 12% trading band around a 30 day m.a. worked pretty well for the NASDAQ until it ran into extreme volatility and spike-up rallies in the winter of 1998-1999, and again in the winter of 1999-2000, when both times it traded mostly outside of the upper limit of the trading band throughout those two rallies.

BOLLINGER BANDS:

Several years ago John Bollinger introduced a different form of trading band which has been dubbed 'Bollinger Bands'. While a normal trading band is plotted at a fixed percentage above and below the moving average, Bollinger Bands are plotted at standard deviation values above and below the moving average. (Standard deviation is a statistical measurement of volatility).

You're really not going to be able to experiment with standard deviations or Bollinger Bands by plotting them manually, but here's the formula: First calculate an x-time frame simple moving average (e.g. a 20 day m.a.) of the index or stock. Then sum the squares of the difference between the index and its m.a. for each of the preceding x-time periods. Divide this sum by x and calculate the square root of the result.

In reality you're going to click on 'Standard Deviations', or 'Bollinger Bands' on the drop-down menu of your charting software and fill in the parameters you want to use. The default parameters for Bollinger Bands on daily closes is to use a 20 day simple m.a. with a trading band plotted 2 standard deviations above and below the m.a., and on weekly closes to use a 20 week simple m.a. with a trading band plotted 2 standard deviations above and below the m.a.

The following chart demonstrates Bollinger Bands on the NASDAQ Composite.

At first glance it looks like Bollinger Bands solve the problem of volatile indexes trading outside the parameters of a normal trading band. However, all that is happening is that through the use of standard deviations to plot the envelope of the band, the band merely expands as needed to encompass the index regardless of its volatility. So in the interpretation, rather than noting that the index has broken outside of the trading band as with a normal trading band, with Bollinger Bands one notes that the band has had to widen to encompass the index.

We admit to struggling to find consistent results using Bollinger Bands, and rarely use them, but the published rules are as follows:

Sharp price changes frequently occur after the bands tighten due to volatility lessening (but with no indication of which direction the prices may move).
When prices move outside the bands, a continuation of the trend is implied.
Bottoms or tops made outside the band, followed by a top or bottom made inside the band, calls for a reversal in the trend.
A move that originates at one band tends to go all the way to the other band.

LIMITATIONS OF TRADING BANDS:

Their consistency diminishes as the volatility of the holding rises. For instance trading bands are very consistent on the Dow, less so on the NASDAQ. They tend to be more consistent with mutual funds than with individual stocks, as mutual funds are diversified and less likely to have unexpected plunges or spike-ups. Individual stocks are subject to unexpected plunges or spike-ups as earnings surprises or disappointments are announced, and the smoothing action of the central moving average prevents it, and the trading band around it, from keeping up with the short term activity.

The following chart of Amgen demonstrates the problem. The upper limit of the trading band provided no help in predicting pull-backs on three sharp spike-ups in 12 months.

However, moving average crossover indicators, like MACD, are very helpful in such situations. And so:

COMING UP:

In the next seminar, we'll discuss moving average crossover systems, in which two or more moving averages are calculated, one slow, the other fast, allowing rallies or corrections to run until an actual sell or buy signal is triggered when one m.a. crosses over the other.

Back to the Top Classroom Home Street Smart Report Home