Part III: Bollinger Bands on Their Own

Why the repeated emphasis on customization and fitting of risk and reward parameters? Because no system, no matter how good it is, will be used if the user isnt comfortable with it. If you do not suit yourself, you will find out quickly that these approaches will not suit you.

"If these methods work so well, why do you teach them?" This is a frequent question, and the answer is always the same. First, I teach because I love to teach. Second, and perhaps most important, because I learn as I teach. In researching and preparing the material for this book, I learned quite a bit, and I learned even more in the process of writing it.

"Will these methods still work after they are published?" The question of continued effectiveness seems troublesome to many, but it is not really. These techniques will remain useful until the market structure changes sufficiently to render them moot. The reason that effectiveness is not destroyed-no matter how widely an approach is taught-is that we are all individuals. If an identical trading system was taught to 100 people, a month later not more than two or three, if that many, would be using it as it was taught. The people would have taken it and modified it to suit their individual tastes, and incorporated it into their unique way of doing things. In short, no matter how specific or declarative a book gets, every reader will walk away from reading it with unique ideas and approaches, and that, as they say, is a good thing.

The greatest myth about Bollinger Bands is that you are supposed to sell at the upper band and buy at the lower band. It can work that way, but it doesnt have to. In Method I well actually buy when the upper band is exceeded and short when the lower band is broken to the downside.1 In Method II well buy on strength as we approach the upper band only if an indicator confirms and sell on weakness as the lower band is approached, again only if confirmed by our indicator. In Method III well buy near the lower band, using a W pattern and an indicator to clarify the setup or well sell near the upper band on a series of tags accompanied by a weakening indicator. Then a variation that relies on noncon-firmed band tags to identify buys and sells will be presented.

Method I, also known as The Squeeze, anticipates high volatility by taking advantage of the cyclical nature of volatility and looking for extremely low volatility as a precursor of high volatility.

Chapter 16: Method I: Volatility Breakout

Now, for Method I. Years ago the late Bruce Babcock of Commodity Traders Consumers Review interviewed me for that publication. After the interview, we chatted for a while, and the interviewing gradually reversed; and it came out that his favorite commodity trading approach was the volatility breakout. I could hardly believe my ears. Here is the fellow who had examined more trading systems-and done so rigorously-than anyone else, with the possible exception of John Hill of Futures Truth, and he was saying that his approach of choice to trading was the volatility-breakout system? The very approach that I thought best for trading after a lot of investigation?

Perhaps the most elegant direct application of Bollinger Bands is a volatility-breakout system. These systems have been around a long time and exist in many varieties and forms. The earliest breakout systems used simple averages of the highs and lows, often shifted up or down a bit. As time went on, average true range was frequently a factor.2

There is no real way of knowing when volatility, as we use it now, was incorporated as a factor, but one would surmise that one day someone noticed that breakout signals worked better when the averages, bands, envelopes, etc., were closer together, and so the volatility-breakout system was born. (Certainly the risk-reward parameters are better aligned when the bands are narrow, a major factor in any system.)

Our version of the venerable volatility-breakout system utilizes BandWidth to set the precondition and then takes a position when a breakout occurs. There are two choices for a stop or exit for this approach. First, Welles Wilders Parabolic,3 a simple but elegant concept. In the case of a stop for a buy signal, the initial stop is set just below the range of the breakout formation and then incremented upward each day the trade is open. Just the opposite is true for a sell. For those willing to pursue larger profits than those afforded by the relatively conservative Parabolic approach, a tag of the opposite band is an excellent exit signal. This allows for corrections along the way and results in longer trades. So, in a buy use a tag of the lower band as an exit, and in a sell use a tag of the upper band as an exit.

The major problem with successfully implementing Method I is a head fake (Figure 16.1)-discussed in the prior chapter. The term came from hockey, but it is familiar in many other arenas

as well. The idea is a player with the puck skates up the ice toward an opponent. As he skates, he turns his head in preparation to pass the defender; as soon as the defenseman commits, he turns his body the other way and safely snaps his shot. Coming out of a Squeeze, stocks often do the same; theyll first feint in the wrong direction and then make the real move. Typically what youll see is a Squeeze, followed by a band tag, followed in turn by the real move. Most often this will occur within the bands and you wont get a breakout signal until after the real move is under way. However, if you have tightened the parameters for the bands as so many who use this approach do, you may find yourself with the occasional small whipsaw before the real trade appears.

Some stocks, indices, etc., are more prone to head fakes than others. Take a look at past Squeezes for the item you are considering and see if they involved head fakes. Once a faker...

For those who are willing to take a nonmechanical approach trading head fakes, the easiest strategy is to wait until a Squeeze