The Bollinger Bands Swing Trading System[1]

May 2004

The Bollinger Bands %b Swing Trading System "...The irony is that when the market exhibits what is widely deemed to be chaotic behavior, the market becomes more predictable..."

1

Copyright © 2004, Connors Group, Inc. ALL RIGHTS RESERVED. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher and the author. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold with the understanding that the author and the publisher are not engaged in rendering legal, accounting, or other professional service. Authorization to photocopy items for internal or personal use, or in the internal or personal use of specific clients, is granted by The Connors Group, Inc., provided that the U.S. 7.00 per page fee is paid directly to Connors Group, Inc., 445 S. Figueroa Street, Suite 2930, Los Angeles CA 90071. Phone: 1-213-955-5858. Printed in the United States of America Bollinger Bands were developed by John Bollinger. This product was not developed by and is not endorsed by John Bollinger or Bollinger Capital Management, Inc. and is not affiliated in any way with BollingerBands.com or BollingerOnBollingerBands.coin.

2

Disclaimer It should not be assumed that the methods, techniques, or indicators presented in this book will be profitable or that they will not result in losses. Past results are not necessarily indicative of future results. Examples in this book are for educational purposes only. The author, publishing firm, and any affiliates assume no responsibility for your trading results. This is not a solicitation of any order to buy or sell.

The NFA requires us to state that "HYPOTHETICAL OR SIMULATED PERFORMANCE RESULTS HAVE CERTAIN INHERENT LIMITATIONS. UNLIKE AN ACTUAL PERFORMANCE RECORD, SIMULATED RESULTS DO NOT REPRESENT ACTUAL TRADING. ALSO, SINCE THE TRADES HAVE NOT ACTUALLY BEEN EXECUTED, THE RESULTS MAY HAVE UNDER- OR OVERCOMPENSATED FOR THE IMPACT, IF ANY, OF CERTAIN MARKET FACTORS, SUCH AS LACK OF LIQUIDITY. SIMULATED TRADING PROGRAMS IN GENERAL ARE ALSO SUBJECT TO THE FACT THAT THEY ARE DESIGNED WITH THE BENEFIT OF HINDSIGHT. NO REPRESENTATION IS BEING MADE THAT ANY ACCOUNT WILL OR IS LIKELY TO ACHIEVE PROFITS OR LOSSES SIMILAR TO THOSE SHOWN."

3

Introduction Welcome to the Bollinger Bands %b Swing Trading System strategy guide. In the pages to come, you will learn how this system works and how you can apply it immediately to your own trading. Plus, you will also learn the system's more aggressive version, the %b Plus One system. Here are some of the highlights of these two systems' features and performance: • The system had 81.65% profitable signals in the S&P 500 Cash Index for both long and short trades between 01/03/1994 and 04/13/2004. It achieved this performance even though this period included one of the worst bear markets in history. • Short signals were over 70% profitable in spite of the fact that the S&P 500 has a strong upward bias. • Gained 1676.99 S&P points... more than double the gains of the S&P 500 cash index for the same period of time. • Gained 4207.07 points for the Nasdaq-100. This was more than 300% the gains of the Nasdaq-100 cash index. • Achieved the above performance by being in the market less than 20% of the time. This means that you could have been in cash more than 80% of the time gaining money market interest, assuming no market risk and having the option of using your cash for other opportunities. • Had a profit factor of 3.66. Systems with profit factors of 2.0 are considered good and systems with profit factors of 2.5 are considered excellent. How does the Bollinger Bands %b Swing Trading System do this? It finds short periods of time in which the market moves too far and too fast in one direction and then stays there for too long. When this occurs, sharp reversals are the usual outcome. The system then is designed to enter those reversals unhesitatingly on either the long- and short-side for trades that typically last from 2 to 5 days.

What Is The Bollinger Bands %b Swing Trading System? The Bollinger Bands %b Swing Trading System (BB %b) finds reversals that occur when a market gets overbought or oversold relative to that market's recent volatility. Volatility simply refers to the range of movement in a market. A stock that typically has 7-point swings within 10 days is more volatile than a stock that only has 2-point swings in 10 days. The way that the BB %b differs from other systems is that it takes into account the fact that the amount of volatility in a market is constantly changing. Therefore the conditions that produce reversals also change. For example a 5-point drop might get stock XYZ oversold enough to produce a reversal. But that might not be true 10 days from now if the level of volatility in XYZ increases. That being the case, reversals can be best anticipated when a market gets overbought or oversold relative to a market's recent behavior.

4

Putting this principle to work has allowed us to create a swing trading system that has had excellent results over the past decade in the major indices and individual stocks. We'll have a lot of more to say about the rules and performance later in this strategy guide. But first, here is a further explanation of the principles that make the Bollinger Bands %b Swing Trading System viable.

How Does The Bollinger Bands %b Swing Trading System Work? History has shown time and time again that the market becomes predictable (within certain limits) when it goes completely out of whack. That is, its behavior dramatically deviates from the norm. When that occurs, statistical analysis shows that the more extremely out of whack a market gets, the higher the probability of its returning to normal behavior. Predictable behavior is the outcome and fortunately, this predictability can be applied to trading. This is an irony that is hidden from most traders in the sense that when the market exhibits what is widely deemed to be chaotic behavior (at least to the news media), the market becomes more predictable! The concept we are describing is known as reversion to the mean and the BB %b Swing Trading System is a highly refined derivative of this principle. One of the many ways in which the market goes out of whack is that it sometimes makes extreme moves either up or down. In other words, it moves much faster and farther for a shorter duration of time than is normally the case. When this occurs, the market has a tendency to adjust itself by reversing back in the opposite direction. Traders and technical analysts call this condition either an overbought or oversold condition. Q: Does an overbought or oversold condition alone give you an edge? The answer is no and many traders have found that out the hard and painful way. An overbought or oversold condition by itself is not a good predictor of market behavior. Something's missing. Our research shows that by adding two important factors to any overbought or oversold condition, you are able to find situations that are predicable. Here are those two factors: FACTOR #1: A market must be overbought or oversold relative to that market's recent volatility. FACTOR #2: The longer the time that FACTOR #1 is in effect, the greater the likelihood is that the overbought or oversold condition will result in a strong reversal.

5

Here's What You'll Learn In The Bollinger Bands %b Swing Trading System Strategy Guide Here's a brief breakdown of the contents of this book: In Part I, you'll learn the four rules of the BB %b for the buy side and the four rules of the system for the short side. We will explain to you the logic behind each of the rules and then show you the hypothetical performance of the system over many years of data, including its 81% winning trades in the S&P 500 Index since 1994. You will also see the test results for the Nasdaq-100 (NDX) and the Semiconductor Index (SOX). All of these can be traded using stocks and Exchange-Traded Funds, which we'll explain later in this Strategy Guide. In Part II, we'll teach you an aggressive version of the system called %b Plus One. For experienced traders who can handle greater levels of risk, this version of the system produced 4207 back-tested points in the Nasdaq-100 while the index itself gained only 1040 points. It accomplished this while being in the market only 18.25% of the time. In Part III, we will give you an overview of the best trading vehicles for trading the BB %b System You will learn how to apply it to Exchange-Traded Funds such as the SPYs, QQQs and SMHs. Plus, we'll give you general guidelines on how to apply the system to individual stocks.

6

In the BB %b Swing Trading System, we are able to combine the two factors above to find overbought or oversold conditions that are so persistent relative to recent market volatility, that a strong reversal is the likely outcome. The system rules that you are taught in Section 11 accomplish the following: For longs: 1. Identifies oversold readings in uptrending markets. 2. Triggers buy signals when those oversold readings persist for a specific period of time. 3. Exits when the market reaches an overbought level For Shorts: 1. Identifies overbought readings in downtrending markets. 2. Triggers sell signals when those overbought readings persist for a specific period of time. 3. Exits when the market reaches an oversold level.

Who Is The Bollinger Bands %b Swing Trading System Designed For? This system is designed for short-term traders and swing traders who want to capture potentially strong reversal moves lasting 2 to 7 days in individual stocks and index trading vehicles such as the SPYs, QQQs and SMHs. To accomplish this, the BB %b allows you to: 1. 2. 3. 4.

Enter powerful short-term reversals just as they begin. Potentially capture the fastest portion of that move. Exit at the most opportune time. Do all of the above over and over again without the need for discretionary or subjective judgment.

The exits in this system are mechanical, but more experienced traders are welcome and encouraged to apply their own exit rules. You have the flexibility to use this system either part- or full time because it does not require that you sit in front of a screen all day. The information that is plugged into the rules we'll be teaching you can be retrieved after the market closes from most data vendor

7

Part I: The Bollinger Band %b Swing Trading System System Logic, Rules And Performance Take calculated risks. That is quite different from being rash. General George S. Patton

System Logic In the introduction, we stated that reversals have a high degree of predictability when the following two factors are in place: FACTOR #1: A market must be overbought or oversold relative to that market's recent volatility. FACTOR #2: The longer the time that FACTOR #1 is in effect, the greater the likelihood is that the overbought or oversold condition will result in a strong reversal. The analogy is that we are not only stretching a rubber band to its limits, but we are also holding it at its most extreme level of stretch for a duration of time. Let's now take these concepts and make a mechanical trading system out of them. Using Bollinger Bands To Determine Recent Volatility

Here, the Bollinger Bands play a significant role. Bollinger Bands are a well-known tool to traders. Created and popularized by professional money manager John Bollinger, the Bollinger Bands are an excellent way to visually display the amount of volatility in a market through use of bands above and below the price bars in a bar chart (John has all excellent website which is http://www.bollingerbands.com. Be sure to spend some time there if you want more in-depth knowledge about this excellent tool.).

8

Take a look at the chart above. Without getting into the statistical concepts such as standard deviation, a simple way to describe what Bollinger Bands do is to say that the distance of the upper and lower bands from the center of the band is based on how volatile the closing prices in recent days have been. The more deviation there is, the wider the bands expand. In the above chart, there is increased volatility going into March and you can see how the bands expand in response to that. This gives you a good visual indication of how much volatility there is in a market. For many traders, the Bollinger Band indicator is a highly useful tool. Usually, it is used in conjunction with other indicators and technical analyses and can provide excellent confirmation that a market has moved too far in one direction and is likely to reverse. The general rule of thumb is that when a market closes near or beyond the upper or lower bands, reversals have a tendency to occur. In general practice, using Bollinger Bands in this fashion requires subjective judgment and multiple confirmations of other indicators and/or patterns. The goal of BB %b is to take this useful tool and systematize it so that it generates mechanical buy and sell signals that do not require subjective judgment. This is where the %b part of the equation enters the picture.

9

Using %b To Quantify Recent Volatility %b tells you where the current close price of a market is in relation to the upper and lower Bollinger Bands. It is measured on a scale of 0 to 1 when the market closes in between the bands. If a market closes right at its lower Bollinger Band, the value of %b is 0. If it closes at the higher band, the value of %b is 1. A close right at the midpoint between the upper and lower bands would give you a %b value of .50. And when markets make extreme moves, the values of the %b can shoot above or below the 0 to 1 range. The general way to think of %b is as follows: The lower the %b is, the more oversold the market is relative to recent history. The higher the %b is, the more overbought the market is relative to recent history.

Note: %b calculation is required on a nightly basis in order to apply the system. It is easy to calculate. The formula is as follows: %b = (Close - Lower Bollinger Band) / (Upper Bollinger Band - Lower Bollinger Band) Even though you can calculate b% by yourself, as an added convenience, the %b numbers will be made available to you free of charge as a registered user of the system. Details will be given to you at the end o this strategy guide

.

10

The Bollinger Band %b Swing Trading System Strategy Rules Now here are the rules of the system. Bollinger Bands Settings:

For all system rules we use 5-period Bollinger Bands with a standard deviation of 1. Rules For Longs: 1) 2) 3) 4)

The market closes above its 200-day Moving Average (MA) The %b closes below 0.2 for 3 days in a row Buy the market at the close (or at the next morning's open) Exit at the close when the %b closes above 0.8

Rules For Shorts: 1) 2) 3) 4)

The market closes below its 200-day MA The %b closes above 0.8 for 3 days in a row Sell the market at the close (or at the next morning's open) Exit at the close when the %b closes below 0.2

Now, let's break the system down into its separate pieces so that you can gain a more indepth understanding of the logic behind the system.

Parameter Setting: We start with a 5-period Bollinger Band with a standard deviation of 1. This is a more aggressive setting than the default value of the Bollinger Band indicator that comes standard with most charting platforms. The default setting is normally a 20-period Bollinger Band with a standard deviation of 2. We use this more aggressive setting so that we can identify short-term reversals. In the following discussion, since the system logic applies the same principles for both longs and shorts, we'll discuss the rules for both simultaneously.

Rule #1: Then, we add a trend filter which forces us to only take trades that are in the direction of the major trend. By trading with the trend, we are improving the chances that we will get winning trades.

11

For longs, we only will buy oversold conditions in an uptrending market. We defined an uptrending market as one where the close is above its 200-day simple moving average. For shorts, we only will short overbought conditions in a downtrending market. We define a downtrending market as one where the close is below its 200-day simple moving average.

Rule #2: Next, we look for %b values that are not only at overdone levels, but which also persist over the course of three days. When we do this, we are taking advantage of the predictability of the following situation: The market has moved too fast and too far in one direction. We use 0.2 for oversold conditions and 0.8 for overbought conditions. Why 0.2 and 0.8? While these numbers appear to be "default values," they are not arbitrary and they are certainly not optimized. We calculated the mean value of %b for the S&P 500 over the decade's test period and found that it was .57. When we adjusted for the upward bias of the market, the oversold value of %b became 0.2 and the overbought value became 0.8. • The market has been in this state for too long a period of time (3 days). • Therefore, the market will need to correct itself by making a reversal.

For both longs and shorts, the system is looking at the current close and comparing it with the average of all the closes for the past 5 days. If the %b reaches the thresholds specified, the market is deemed to have moved too far in too short a period of time. The stage is set for a reversal.

Rule #3: We enter a trade. Once all the preceding conditions are met, there's a good chance that a fast, sharp reversal will occur. So it's a good time to enter, long or short depending on what kind of signal it is. You can either enter at the close of the day the signal is generated or at the open of the next day of trading. For example, if a system signal is generated at the end of a trading day, you can enter the market in the last 15 minutes of trading or at the beginning of the next day's trading.

12

Rule #4:

We exit the trade. Once reversals from overbought or oversold conditions occur, they tend to very quickly get "overdone" in the opposite direction. Therefore: Once a long trade goes into overbought territory (%b is .80) we exit. Once a short trade goes into oversold territory (%b is .20) we exit.

This system is designed to have you enter and exit during the "sweet spot" of a reversal move. That is when the probability of success is highest. This means that following the system in a disciplined manner will have you leaving money on the table. In many cases, the market will continue to move forcefully in the direction of the trade, well after you've exited. That is a necessary feature of the system in order to achieve its high percentage of winning trades. Remember this: Maintaining discipline on your exits is critical. There is only one sort of discipline, perfect discipline. General George S. Patton

13

All Bollinger Band %b Swing Trading System Buy And Sell Signals For The S&P 500 January 1994 to March 2004 In a short while, we'll get to the backtested results from over a decade of testing that you need in order to fairly evaluate this system's performance. But before we get to those numbers, let's take a look at some charts that reveal a great deal about the strength and consistency of the system. The following charts show you all of the entry and exit points for signals that were generated in the S&P 500 between January 1994 and March 2004. The labels on the charts are fairly self-explanatory, but to make sure you understand here is what they mean:

Buy = Go long one unit. 1 0 LX = Exit long position.

-1 Sell = Go short one unit.

SX = Exit short position 0

14

S&P 500 Index: March 1994 to December 1994

15

S&P 500 Index: January 1995 to October 1995

16

S&P 500 Index: November 1995 to August 1996

17

S&P 500 Index: August 1996 to May 1997

18

S&P 500 Index: July 1997 to April 1998

19

S&P 500 Index: May 1998 to February 1999

20

S&P 500 Index: March 1999 to December 1999

21

S&P 500 Index: January 2000 to November 2000

22

S&P 500 Index: November 2000 to September 2001

23

S&P 500 Index: October 2001 to July 2002

24

S&P 500 Index: August 2002 to May 2003

25

S&P 500 Index: June 2003 to March 2004

26

Comments About The Charts One of the observations that comes out of these charts is that the Bollinger Bands %b Swing Trading System is good at identifying major short-term reversals. Indeed, as you see in the forthcoming backtested performance tables, it has been correct over 81 % on longs and shorts combined on the S&P 500 Index. We'll have a more extensive discussion of this and other performance statistics in the Summary and Results section later on. But what we want to point out here is something that we know many traders will notice about the exits the system makes. The observation is that the system often exits well before the end of the move has occurred. For example: On 07/31/00, the system triggered a buy signal. It exited 3 trading days later on 08/03/00 with a gain of about 22 S&P points. The move continued for another 80 S&P points before finally topping out in early September. It is not hard to find other examples like this. But you can also find many examples in which the market reversed very shortly after the system's proper exit point. It is wise to continually remind yourself that the system is designed to exit longs when the market becomes overbought and exit shorts when the market becomes oversold. Over the course of time, this has tended to yield the best overall performance. We highly recommend that anyone who uses this system stick with the exit rules as they have been spelled out. That being said, experienced traders who are cognizant of the risk factors can consider using other exit methods in order to exploit moves that extend well beyond the "official" exit points that are dictated by the rules of this system.

27

The Results Of The Bollinger Bands %b Swing Trading System In each of the following sections, we will be showing you the Strategy Performance Reports for the Bollinger Bands %b Swing Trading System in the following markets. 1. S&P 500 (SPX) 2. Nasdaq 100 (NDX) 3. Semiconductor Index (SOX) We will go through each market in the sequential order shown above. The following are the test results for the various markets. The first charts also have an explanation of what each column means.

28

TradeStation Strategy Performance Report For The S&P 500 Index TradeStation Strategy Performance Report - PercentB Test SPX.CSV-Daily (1/3194-4/13/04) Performance Summary: A l l Trades

Total Net Profit 1 Gross Profit2

1041.73 1433.34

Total # of trades3 Number winning trades4 Largest winning trades Average winning trade6 Ratio avg win/avg loss?

109 89

Profit Factor10 Gross Loss 11

3.66 -391.61

Percent profitable 12 Number losing trades13

81.65% 20

62.00 Largest losing trade 14 16.10 Average losing trade 15 .82 Avg trade (win & loss)16

Max consec. Winners 8 Avg # bars in winners 9

15 Max consec. losers 17 3 Avg # bars in losers 18

-70.54 -19.58 9.56 2 8

Performance Summary: Long Trades

Total Net Profit Gross Profit Total # of trades Number winning trades

892.74 Profit Factor 988.39 Gross Loss 75 Percent profitable 65 Number losing trades

Largest winning trade Average winning trade Ratio avg win/avg loss

62.00 Largest losing trade 15.21 Average losing trade 1.59 Avg trade (win & loss)

Max consec. Winners Avg # bars in winners

18 Max consec. losers 3 Avg # bars in losers

10.33 -95.65 86.67% 10 -33.78 -9.56 11.90 2 7

Performance Summary: Short Trades

Total Net Profit Gross Profit Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg win/avg loss Max consec. Winners Avg # bars in winners

148.99 Profit Factor 444.95 Gross Loss

1.50 -295.96

34 Percent profitable 24 Number losing trades

70.59% 10

50.95 Largest losing trade 18.54 Average losing trade .63 Avg trade (win & loss)

-70.54 -29.60 4.38

4 2

Max consec. losers Avg # bars in losers

1 9

Percentage of time in the market 01/03/94 through 04/13/04: 18.70%19

29

Definitions This will not be repeated in future tables, so use this table as a reference. 1. 2.

Total Net Profit: Total hypothetical points gained for all trades during the test period. Gross Profit: Total hypothetical points gained for winning trades only during the test period. 3. Total # of trades Total number of trades (both winning and losing) during the test period. 4. Number winning trades: Total number of trades that generated point gains during the test period. 5. Largest winning trade: Number of points gained in the trade with the greatest point gain. 6. Average winning trade: Gross Profit divided by total number of trades 7. Ratio avg win/avg loss: Average winning trades divided by the average losing trades. 8. Max consec. winners: Maximum of number of consecutive trades that generated point gains. 9. Ave #bars in winners: Average number of trading days from entry to exit for all winning trades during test period, 10. Profit Factor: Gross profit divided by gross loss. 11. Gross Loss: Total hypothetical points lost for losing trades only during the test period. 12. Percent profitable: Percentage of profitable trades out of total number of trades. 13. Number losing trades: Total number of trades that generated point losses during the test period. 14. Largest losing trade: Number of points lost in the trade with the greatest point loss. 15. Average losing trade: Gross Loss divided by total number of trades. 16. Avg trade (win & loss): Total net profit divided by total number of trades. 17. Max consec. losers: Maximum number of consecutive losing trades. 18. Ave # bars in losers: Average number of trading days from entry to exit for all losing trades during test period. 19. Percentage of time in the market: The percentage of time the system was in the market during the test period.

30

S&P 500 Strategy Performance Report Annual Trading Summary Period1 04 YTD5 03 02 01 00 99 98 97 96 95 94

Net Profit2 26.54 129.80 26.50 49.48 177.56 290.38 66.80 183.50 54.91 31.40 4.86

# Trades3 4 11 11 14 12 15 8 11 13 8 3

% Profitable4 75.00% 90.91% 63.64% 71.43% 75.00% 86.67% 87.50% 81.82% 84.62% 100.00% 66.67%

Definitions This will not be repeated in future tables, so use this table as a reference.

1. 2.

3. 4. 5.

Year. The year of the signals — 03 = 2003. Net Profit Total net profits hypothetically gained (lost) for that year. Does not include slippage and commission for both entries and exits on each individual trade. # Trades The number of trades that the Bollinger Bands %b Swing Trading System triggered for that year. % Profitable: The percentage of Bollinger Bands %b Swing Trading System trades that were profitable for that year. 04 YTD: January 1, 2004 through April 13, 2004

31

This equity curve is based upon point gains, not actual trades. Slippage and commission are not included. This applies to all equity curves shown in this Strategy Guide.

32

Strategy Performance Report For The Nasdaq-100 TradeStation Strategy Performance Report - PercentB Test NDX.CSV-Daily (1/3/94-4/12/04)

Performance Summary: All TradesTotal Net Profit Gross Profit Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg win/avg loss Max consec. Winners Avg # bars in winners

2836.46 3485.68

Profit Factor Gross Loss

5.37 -649.22


78.43% 22


-114.85 -29.51

13 3

Max consec. losers Avg # bars in I osers

3 7

Performance Summary: Long Trades Total Net Profit Gross Profit Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg win/avg loss Max consec. Winners Avg # bars in winners


13.03 -194.66


85.33% 11


-78.57 -17.70 31.22


3 7

Performance Summary: Short Trades Total Net Profit Gross Profit Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg winlavg loss Max consec. Winners Avg # bars in winners


2.09 -454.56


59.26% 11


-114.85 -41.32 18.32


2 7

Percentage of time in the market 01/03/94 through 04/13/04: 18.25%

33

Nasdaq-100 Strategy Performance Report Annual Trading Summary

Period 04 YTD 03 02 01 00 99 98 97 96 95 94

Net Profit 27.32 258.18 128.89 -49.32 1503.52 483.61 27.23 266.71 23.94 141.85 24.53

# Trades 3 11 10 11 7 10 10 15 9 14 3

% Profitable 100.00% 90.91% 70.00% 54.55% 100.00% 80.00% 60.00% 80.00% 55.56% 92.86% 100.00%

34

Semiconductor Index Strategy Performance Report TradeStation Strategy Performance Report - PercentB Test SOX.CSV-Daily (5/4/94-4/12/04)

Performance Summary: A l l Trades Total Net Profit Gross Profit Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg win/avg loss Max consec. Winners Avg # bars in winners


3.01 -525.45


69.44% 33


-118.74 -15.92 9.77

10 Max consec. Losers 3 Avg # bars in losers

3 7



4.12 -246.00


75.41% 15


-118.74 -16.40 12.59


3 7

Performance Summary: Short Trades Total Net Profit Gross Profit Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg winlavg loss Max consec. Winners Avg # bars in winners


2.03 -279.45


61.70% 18


-48.16 -15.53 6.11


2 8


35

Semiconductor Index Strategy Performance Report Annual Trading Summary Period 04 YTD 03 02 01 00 99 98 97 96 95 94

Net Profit 3.73 104.11 -51.92 182.50 606.98 144.23 1.27 6.32 -26.90 85.26 0.00

# Trades 3 14 10 14 14 10 13 8 10 12 0

% Profitable 33.33% 64.29% 50.00% 64.29% 92.86% 80.00% 61.54% 75.00% 50.00% 91.67% N/A

36

Summary Of The Results First, what stands out is that the percentage of the winning trades is exceptional. For the S&P 500, 81.65% of the signals were profitable. Even more amazing is that this performance is for both longs and short trades combined. Long trades alone for the S&P 500 were 86.67% profitable. On the short side they were 70.59% and that is very respectable considering the strong upside bias that the S&P 500 has. The point gains from the system were 1041.73. That's more than 150% of the 659.76 points that S&P 500 cash index gained. Plus, the system was able to make these gains while being in the market only 18.70% of the time. This means that you could have been in cash more than 80% of the time gaining money market interest, assuming no market risk and having the option of using your cash for other opportunities. The system also provided substantial trading opportunities. Where there were good shortterm moves in the market, the system, for the most part, caught them. There were 109 trades over a period of just over 10 years (123 months) and this gave you the opportunity to be in the market nearly 9 times a year both on the long and the short side. Some years there were more trades and others, fewer. If you look back at the charts provided earlier in this section, you'll see that the system was able to capture good portions of most of the major swings of each year. Many traders will find this comforting from a psychological perspective. And perhaps the most staggering number of all is the profit factor. Systems with profit factors of 2.0 are considered good and systems with profit factors of 2.5 are considered excellent. In the S&P 500, the Bollinger Bands %b Swing Trading System has shown a profit factor of 3.66. Plus, look at the profit factor for longs. It was 10.33! The equity curves pretty much sums up the performance of the system in explicit terms. Both the S&P 500 and Nasdaq-100 equity curves show consistent growth throughout the entire 10-year test period, including the 2000 – 2002 bear market. The SOX tended to perform best during bull markets. The Bollinger Bands %b Swing Trading System may very well put to rest the school of thought that questions the value of mechanized trading systems. It has done an excellent job of capturing the "sweet spot" portions of many of the best swings of each year that was tested, while remaining in cash the majority of the time. It had a high percentage of winning trades earned—150% the number of points that the S&P 500 cash index had gained.

37

Part Il: Improving Performance With The %b Plus One Swing Trading System Traders have long known that any market that becomes overbought or oversold can become even more so. In this section, we look to further increase our edge and our system performance when this occurs. Let's start with the base we have built so far. As you have seen, with the standard Bollinger Bands %b Swing Trading System, we're looking for markets that have: 1. Gone too far and too fast in one direction relative to recent market volatility. 2. Stayed in that state for too long a time. Our premise is that the overbought or oversold condition that results becomes unsustainable and must be unwound through a reversal. This gives us a trading opportunity with good risk to reward characteristics. The performance results we've shown you for the Nasdaq-100, S&P 500 and the Semiconductor Index support that premise. Now let's take it one step further by asking the following question: Is there a good way to take advantage of situations in which the market becomes even more overbought or oversold? The answer is yes, and this has allowed us to create an even more aggressive version of the system, which we call the %b Plus One Swing Trading System. The Rules Of The %b Plus One Swing Trading System Let take a look at the rules and then we'll go into a quick discussion of how this potentially boosts the performance of the "standard" version of the system.. Bollinger Bands Setting: As with the standard BB %b system we use 5-period Bollinger Bands with a standard deviation of 1. Rules For Longs: 1) 2) 3) 4) 5)

The market closes above its 200-day MA The %b closes below 0.2 for 3 days in a row Buy the market at the close (or at the next morning's open) Buy an additional unit when the %b closes below 0.2 for 4 days in a row Exit at the close when the %b closes above 0.8

38

Rules For Shorts: 1) 2) 3) 4) 5)

The market closes below its 200-day MA The %b closes above 0.8 for 3 days in a row Sell the market at the close (or at the next morning's open) Sell an additional unit when the %b closes above 0.8 for 4 days in a row Exit at the close when the %b closes below 0.2

As you can see, the main difference between the %b Plus One and the original system is that we're increasing our position if the oversold or overbought condition persists for 4 consecutive days. Already, by the time conditions of the first three rules are met, the market is extremely overstretched. Now, if the market spends an additional day in that condition, we are saying that has become even more overstretched and its ability to stay in that condition is about to be exhausted. We increase our position size in anticipation that: 1. A more powerful reversal than is normally the case will occur. 2. And possibly, that reversal has a much higher than normal probability of occurring. We subjected the %b Plus One Swing Trading System to the same testing as the original version. Here are those test results:

39

S&P 500 Strategy Performance Report TradeStation Strategy Performance Report - PercentB Test SPX.CSV-Daily (1/3/94-4/13/04) Performance Summary: All Trades

Total Net Profit Gross Profit Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg win/avg loss . Max consec. Winners Avg # bars in winners

1676.99 2267.11


3.84 -590.12

Percent profitable Number losing trades

81.29% 32

62.00 Largest losing trade 16.31 Average losing trade 88 Avg trade (win & loss)

-70.54 -18.44 9.81

171 139


3 8


Total Net Profit Gross Profit Total # of trades Number winning trades

1368.99 1517.74


10.20 -148.75


86.84% 15


62.00 Largest losing trade 15-33 Average losing trade 1.55 Avg trade (win & loss)

-33.78 -9.92 12.01



3 7


Total Net Profit Gross Profit Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg win/avg loss . Max consec. Winners Avg # bars in winners

308.00 749.37


1.70 -441.37


70.18% 17

50.95 Largest losing trade 18.73 Average losing trade 72 Avg trade (win & loss)

-70.54 -25.96 5.40

57 40

8 Max consec, losers 2 Avg # bars in losers

2 9


40

S&P 500 Strategy Performance Report Annual Trading Summary Period 04 YTD 03 02 01 00 99 98 97 96 95 94

Net Profit 49.33 154.40 171.82 56.85 271.79 468.65 68.27 263.02 126.19 35.86 10.81

# Trades 6 15 19 23 20 22 13 17 22 9 6

% Profitable 83.33% 86.67% 73.68% 65.22% 70.00% 90.91% 84.62% 88.24% 86.36% 100.00% 66.67%

41

Nasdaq-100 Strategy Performance Report TradeStation Strategy Performance Report - PercentB Test NDX.CSV-Daily (1/3/94-4/12/04)

Performance Summary: All Trades Total Net Profit Gross Profit


Total # of trades Number winning trades





6.41 -777.67 81.29% 29 -114.85 -26.82 27.14


4 7


Total Net Profit Gross Profit


Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg win/avg loss

16.73 -208.42


88.39% 13


-78.57 -16.03 $29.28



3 7


Total Net Profit Gross Profit Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg win/avg loss Max consec. Winners Avg # bars in winners


2.63 -569.25


62.79% 16


-114.85 -35.58 21.58

8 3

Maxconsec.losers Avg # bars in losers

3 7


42

Nasdaq-100 Strategy Performance Report Annual Trading Summary Period 04 YTD 03 02 01 00 99 98 97 96 95 94

Net Profit 74.96 362.88 253.78 269.46 665.52 721.09 96.72 427.77 68.78 238.56 27.55

# Trades 6 15 16 20 8 15 15 21 15 21 4

% Profitable 100.00% 93.33% 75.00% 60.00% 100.00% 86.67% 60.00% 85.71% 66.67% 95.24% 100.00%

43

Semiconductor Index Strategy Performance Report

TradeStation Strategy Performance Report - PercentB Test SOX.CSV-Daily (5/4/94-4/12/04)

Performance Summary: All Trades Total Net Profit Gross Profit Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg win/avg loss Max consec. Winners Avg # bars in winners


2.74 880.16

167 116


69.46% 51

160.43 20.76 1.20

Largest losing trade Average losing trade Avg trade (win & loss)

-118.74 -17.26 9.15

13 3


6 7


1019.00 1478.05


3.22 -459.05

92 69


75.00% 23

160.43 21.42 1.07


-118.74 -19-96 11.08

13 3


4 7

Performance Summary: Short Trades Total Net Profit Gross Profit Total # of trades Number winning trades Largest winning trade Average winning trade Ratio avg win/avg loss Max consec. Winners Avg # bars in winners

508.77 929.88


2.21 -421.11

75 47


62.67% 28

112.17 9.78 1.32


-48.16 -15.04 6.78

9 2

Max consec. Losers Avg # bars in losers

4 8


44

TradeStation Strategy Performance Report Annual Trading Summary Period 04 YTD 03 02 01 00 99 98 97 96 95 94

Net Profit 36.22 152.43 -33.68 340.72 689.54 279.03 -12.94 7.19 -47.22 116.48 0.00

# Trades 5 22 16 24 20 15 18 13 17 17 0

% Profitable 60.00% 72.73% 50.00% 66.67% 85.00% 86.67% 61.11% 76.92% 41.18% 88.24% N/A

45

Summary Of The Results Looking at the results of %b Plus One, you can see that the performance is even better than with the less aggressive Bollinger Bands %b Swing Trading System. To start with, look{ at the equity curves of the S&P 500 and the Nasdaq-100. They are steeper and smoother than the equity curves for the less aggressive version of the system One of the pronounced contrasts is that the slope remains fairly steep through the 2000 — 2002 bear market. That was less the case in the equity curves generated from the standard system. The bottom line is that there appeared to be fast and more consistent equity growth with this aggressive version of the system Now let's turn our attention to the performance numbers for they are truly superb. The percentage of winning trades continues to be high. For the S&P 500, the %b Plus One had 81.29% winning trades for both longs and shorts. Now here is the kicker. The number of points gained by the S&P 500 cash index was 659.76. During that same period, the system gained nearly 250% of that amount by gaining 1676.99 points. The Nasdaq-100 numbers are even better. The Nasdaq-100 cash index gained 1040.13 points, but for the same period of time tested, the %b Plus One was up an astounding 4207.07 points ...over 300% the gains of the Nasdaq-100! This is particularly amazing when you consider that the amount of time that the system was in the market was still small---only 18.70%. Again, you were able to stand in cash over 80% of the time and either earn interest or diversify into other investment vehicles. Think about it...this is no more time exposure than with the standard version of the system and yet the performance is substantially greater. Finally, take a look at the profit factor. The %b Plus One has a remarkably high profit factor of 3.84 for the S&P 500 and the Nasdaq-100 profit factor is even higher, coming in at 6.41. As we stated before, a profit factor of 2.0 is considered to be exceptional. In conclusion, the %b Plus One version of the system is more aggressive because it is increasing the position size when a market becomes more persistently oversold. Yet it is not in the market for any longer than the standard version of the system. To be sure, this system incurs greater risk exposure due to the larger position size, but for experienced traders who understand the risks, it is an attractive alternative to the Bollinger Bands %b Swing Trading System.

46

Part III: Trading Vehicles You Can Use To Trade The Bollinger Bands %b Swing Trading System There are two main ways to take advantage of the trading opportunities to which you will potentially be alerted through BB %b Swing Trading System. They are as follows: 1. Exchange-Traded Funds (ETFs) 2. Stocks Let's now discuss these in detail. Exchange-Traded Funds ETFs are baskets of stocks that are traded like regular stocks. They are extremely liquid and trade millions of shares a day. We believe they are the simplest way to take advantage of the system. ETFs can be traded through any broker. There are many different ETFs. Some are designed to move in tandem with specific industry groups. Others move with foreign stock markets. But for our purposes, we will focus on ETFs that move in tandem with stock market indices. In this strategy guide, the testing we have done of different markets shows that the Bollinger Bands %b Swing Trading System works best on the S&P 500 Index, Nasdaq-100 Index, and the Semiconductor Index. To take advantage of the signals on these different indices you would trade the following ETFs: • SPY to trade the S&P 500 Index • QQQ to trade the Nasdaq 100 • SMH to trade the Semiconductor Index

Stocks The way in which the rules of the Bollinger Bands %b Swing Trading System are applied to stocks is identical to the way they are applied to ETFs. Performance results can either exceed or fall behind the test results we have shown you in this strategy guide. You are free to conduct your own testing to find out which stocks you are most comfortable with. In considering which stocks you apply this system to, we recommend that you start out with the component stocks of the S&P 500, Nasdaq-100 and the SMH. These stocks are more likely to perform in line with the test results you see in this strategy guide, although we cannot guarantee it.

47

Closing Thoughts Discretionary traders have long questioned whether it is possible to trade successfully for a long period of time with a mechanical system. We will not get involved in that debate. We really don't have to because the performance numbers shown in this strategy guide speak for themselves. They are impressive. While no guarantee can be made that future performance will be comparable, it is important to note that: This is not a system that has been optimized. It triggers signals using simple rules and simple parameters. If you know something about system development, then you know that a system with these characteristics tends to be robust over time. It is the complex systems with multiple conditions and complex criteria that tend to blow up. Now, let's take a look at what this set of systems has to offer: You don't have to stare at a monitor all day looking for setups. The system does it all for you and generates its signals automatically at the end of each trading day. You can act upon these signals at the next morning's open. Traders who have the necessary tools can also look for the signals in the final 30 minutes of trade. This can give you a small edge on the timing of entries, but this is not mandatory. It takes advantage of both the long and short side. When the market is going up, the system grabs the best opportunities to the upside. And when the market is going down, you aren't sitting on the sidelines. The system allows you continue participating in the strong downside moves. Look at the S&P 500 and Nasdaq — 100 performances during the 2000 to 2002 bear market. Both indices show steady growth. And that growth is even more rapid during the bear market using the %b Plus One system. Your cash is not tied up. We stated earlier that the market is chaotic and unpredictable the vast majority of the time. The only time that this system jumps into the market is when it detects a predictable reversal. Think about it. Isn't it more prudent to be in the market only during those times when you have a statistical edge, as opposed to trading every big move indiscriminately? When the odds are not favorable, this system makes sure you are in cash. It captures large chunks of the best moves. While the system is picky about when it allows you to enter the market, it has tended to aggressively enter at major highs and lows. You can see that clearly in the S&P 500 charts we showed you earlier in this strategy guide. It tends to not miss the beginnings of major moves. It's mechanical and doesn't require you to apply discretionary judgment. There is nothing wrong with applying discretionary judgment in making trading decisions. The problem is that there are many psychological factors in trading that interfere with the quality of the decision-making that takes place in the heat of battle. Using our system the mechanical way that it was designed to be used

48

allows you to be an active short-term trader without the interference of emotions and faulty judgment. High percentage of winning trades. One of the most common challenges in trading is having the discipline to take small losses before they get big or out of control. If you have this difficulty, then the Bollinger Bands %b Swing Trading System may be of some help. Obviously, like any trading approach, it does have losing trades. But, in the past, it has provided a percentage of winning trades that many traders will find less challenging to tolerate than is more often the case with the most widely available strategies and methodologies. We obviously have to emphasize that there is no guarantee this high percentage of winning trades will continue, but we'll just let the numbers speak for themselves. We hope you have gained an in-depth understanding of the Bollinger Bands %b Swing Trading System and the %b Plus One Swing Trading System. This system has provided a level of performance that is unprecedented for strategies and systems that operate in the swing trading time frame. The rules are simple and easy to apply. But we recommend that you go slow and watch its performance in the current markets to get a feel for what it can do in different market environments.

Best of luck!

49

The Bollinger Bands Swing Trading System[1]

Recommend Documents