Option Investor
Trader's Corner

Bollinger Bands

Printer friendly version

Those of you who follow my INDEX TRADER articles written on the weekend know that I tend to consistently use certain indicators or technical 'studies' to suggest where the market is or is probably getting into an 'overbought' or 'oversold' condition; i.e., where prices may be getting 'extended'.

Mostly I use the "moving average envelope" study for the stock indexes, where two lines form percent 'envelope' lines above and below a particular moving average of the daily close. I use a 21-day moving average and begin with two envelope lines that represent a value (on any given day) equal to 3 percent above and 3 percent below the S&P and Dow and envelope lines generally that are a bit wider (e.g., 3.5-4.0 percent) for the Nasdaq.

Often I adjust the percentage such that the upper line and lower lines are a half percent different simply based on where a particular index seems to be hitting resistance or where it is finding support over the most recent half to full year period. By the way, where bottoms or tops tend to form in individual stocks, in terms of percent moving average envelopes tend to vary so much that I don't generally use moving average envelope lines.

However, on stocks you can always use the BOLLINGER BAND study as it is not necessary to adjust the percent lines as the Bollinger Band indicator does it for you, which I'll explain. Moreover, there are times when price action relative to the Bollinger Band study is an effective tool to suggest that a top or bottom has been seen. And, hey, what else do you or I generally care about but where a trend reversal has occurred! I especially am always looking to get into calls or puts BEFORE it's completely obvious that we're seeing an intermediate top or bottom.

I want to give Bollinger Bands more play in this Trader's Corner article. John is someone I know personally and professionally from years past. He plies his trade from a coastal California town that I'm fond of too!


I'm going to diverge one little bit first and show the S&P 100 (OEX) chart with the regular moving average envelope lines of 2.5 and 3.0 percent, relative to the 21-day moving average, but want to show the OEX return to a previously broken up trendline that I thought might mark or come close to marking a tradable top. Then, I'll progress to showing index and stock charts with the Bollinger Band study applied to them, with price 'SPIKES' above or below the upper or lower Bolli bands that have been quite good at identifying tops or bottoms.

This first chart, of the OEX, has my 'regular' moving average envelopes lines above and below the centered (21-day) Simple Moving Average (SMA), equal to a 2.5 percent red line above the SMA and a 3 green percent line BELOW the SMA. Note the places where price pullbacks appeared to find support either at the lower envelope line or at the 21-dya moving average. The lower envelope line tends to be where prices have fallen into an 'oversold' zone. In a strong uptrend, the upper envelope line is LESS EFFECTIVE in identifying a price zone that will become a 'final' top only an area where prices have BEGUN to get into an 'overbought' zone or have begun to be rather far 'extended' above the center moving average relative to a 'normal' (i.e., trading range) market:

The red arrow at Monday and today's (Wed.) high marks finally a return to a previously broken up trendline dating from almost a year ago and which I sometimes call the 'kiss of death' trendline, a term I first heard used by Michael Jenkins. I've thought for some time that a return to this line would suggest an area of strong resistance or selling pressures. I went back and modified the trendline today to see if I really had the best representation of the trendline and adjusted it a bit, as I'll note next.

I've also been noting over this month the other harbinger of tradable tops that we tend to see WHEN bullish sentiment gets to certain 'extremes', as seen by my "CPRATIO" above which is under the OEX price chart.

Because trendlines and the way you draw them is so useful of an art, I'm pausing briefly to show the small adjustment I made to the up trendline seen on the S&P 100 (OEX) chart below. Construction of an 'INTERNAL' trendline is often the most effective way to draw a trendline. An internal trendline is simply one that is drawn through the GREATEST NUMBER of lows or highs; in the case of the up trendline seen on the OEX chart below, I took out ONE low that fell well under the line established by FOUR lows, as noted at the blue up arrows. If the OEX highs this week do turn out to form at least an interim top, that top intersected exactly at resistance implied by the previously broken up trendline.

What I am not explaining with what I've outlined above is WHY the line in question on the OEX chart might in fact be an area of significant resistance, which is another subject for another time. I'll just say here that support, once broken, tends to 'become' resistance later on and that includes trendline support; the reverse is also true: resistance, once broken tends to 'become' support on subsequent pullbacks.

Before moving on to Bollinger Bands, I'll note again that the use of regular moving average ENVELOPES does NOT tend to pinpoint trend reversal points in strong bull or bear moves. In the strong bull market we've been in, prices have tended to soar to the upper envelope line and then tend to just keep going UP along or UNDER that line.


Bollinger Bands (BB) combine a centered moving average, which is part of the indicator but is usually NOT shown, unlike MY next chart below. BB basically combine the moving average envelope technique with a measurement of recent price volatility to determine the optimal placement of the upper and lower lines. The purpose of the Bollinger Band indicator is basically the same as moving average envelopes: are current price levels high OR low on a 'RELATIVE' basis?

Just as with (moving average) envelopes, two 'bands' and the convention is to call these lines bands to distinguish from the fixed percentage envelope technique, are placed above and below a centered moving average, which is usually set to a default setting (by the charting software) of 20 days. However, unlike lines that are a fixed percent above or below the moving average, Bollinger bands are plotted two 'STANDARD DEVIATIONS" above and below the average.

John Bollinger indicated that the defaults of a 20-day moving average and use of TWO standard deviations was derived from his studies of the US stock market using daily data. John indicated that his main use of his bands was for pattern recognition and trend analysis. Occasionally he would use his study or indicator on hourly charts for trade execution or on monthly charts for a long-term perspective.

His indicator has a purpose, as noted, to answer the question: "Are prices high or low on a relative basis?" By definition, prices are 'high' at the upper band and 'low' at the lower band.

The BB study is used in conjunction with other indicators to 'confirm' what they show or as 'confirmation' of a chart pattern; e.g., a key downside reversal. Today's action in SPX as seen above, is suggestive of a key downside reversal; i.e., a move to a new high, followed by a close below the prior 1-2 day's Close or the prior 1-2 day's Low.


Bollinger Bands are used, among other things, to:
1. detect the beginning and ends of trends
2. pick out (trend) reversals
3. diagnose continuation patterns
4. identify overbought or oversold levels
5. clarify chart patterns; e.g., a Head & Shoulder's pattern has a 'typical' BB 'signature'

A broad Head and Shoulder's (H&S) bottom was seen in the hourly Dow 30 (INDU) chart pattern over March, into early-April, as outlined below. The use of the Bollinger Band study on the chart helped highlight the distinct 3-bottom pattern of an H&S bottom. Once this pattern had formed it was not necessary to wait until their was a 'breakout' above the H&S neckline. Instead, an advantageous purchase of DJX calls could have been made on the pullback to the upper region of the 'right shoulder' in the second week of April, setting an exiting stop a bit below (e.g., at 12400) the lower Bollinger Band on the hourly chart as seen below.

Going back to point number 2 in the principle uses of Bollinger Bands noted above, related to picking out possible trend REVERSALS: I find this aspect to be one of the most useful for this indicator; examples are highlighted in the Nas 100 (NDX) chart below.

Tops or bottoms often are seen in the indexes and to some degree in bellwether stocks, when there are intraday 'spikes' significantly above or below the bands. Such price action often comes at the tail end of price swings and are similar to exhaustion moves or gaps; e.g., a last 'gasp' of a bull or bear move, which relates to the 'excesses' often seen when a price move is well along or mature.

I indicated already that moving average envelopes tend to be useful (at 3-4 percent) for judging if one of the major stock indexes is in an overbought or oversold area, but is not very useful when applied to an individual stock charts as there are such variations in the relevant percentage settings.

Because Bollinger Bands adjust for the volatility of any particular stock, they ARE quite useful for individual equities and numerous indicator adjustments (to the indicator 'inputs') are not needed.


Volatility and trend already have been deployed in the construction of Bolli Bands, so there use for the confirmation of price action is not recommended or useful.

The indicators that you use for 'confirmation' of something suggested by Bolli Bands shouldn't be directly related to one another. Two indicators from the same 'category' of indicators do not INCREASE confirmation; e.g., use of the RSI or Relative Strength Indicator AND Stochastics AND MACD do not increase confirmation of an overbought or oversold situation or a bullish or bearish price/oscillator 'divergence'.

As with moving average envelopes, prices do and will 'walk' up the upper Bollinger Band and walk down the lower BB.

Closes outside the Bollinger Bands, especially (and mostly) in individual stocks or in commodities markets can be 'continuation' signals, not reversal 'signals'; e.g., in volatility-breakout systems.

The use of default parameters of 20 periods for a moving average and two standard deviations for the bandwidth are only 'default' settings. The actual parameters needed for any given market or task may NOT be the same as the default settings. Settings are by definition changeable by the user.

The average used is should NOT be the best for 'crossover' type signals, but be appropriate to highlight the intermediate-term trend.

If the average is lengthened, the number of standard deviations needs to also be increased; e.g. from 2 at 20 periods to 2.1 if using a 50-period moving average. OR if the average is shortened, the standard deviation should also be shortened; e.g., from 2 at a 20-period moving average to 1.9 at 10 periods.

Bolli Bands are based on a Simple Moving Average (SMA) because an SMA is used in the standard deviation calculation and this makes for logical consistency.

Lastly, and of course, a 'tag' of the bands are just tags, not reversal 'signals'. A tag of the upper BB is NOT in and of itself a sell signal. A touch to the lower BB is NOT in and of itself an indication that it's time to buy.

As you can see I don't have any good ideas of my own that a Subscriber doesn't put in my head! Please send any technical and Index-related questions for answer in any of my articles to Click here to email Leigh Stevens Support [at] OptionInvestor.com with 'Leigh Stevens' in the Subject line.


Trader's Corner Archives