Japanese Candlestick Charting Techniques PDF
Japanese Candlestick Charting Techniques PDF
Japanese Candlestick Charting Techniques PDF
CHARTING TECHNIQUES
"Candles Exhaust Themselves to Give Light to Men"
JAPANESE
CANDLESTICK
CHARTING
TECHNIQUES
Contemporary Guide to the Ancient
Investment Techniques of the Far East
STEVE NISON
Nison, Steve.
Japanese candlestick charting techniques : a contemporary guide to
the ancient investment technique of the Far East Steve Nison.
p. cm.
Includes bibliographical references and index.
ISBN 0-13-931650-7
1. Stocks-Charts, diagrams, etc. 2. Investment analysis.
I. Title.
1991 90-22736
by Steve Nison
All rights reserved. No part of this book may be reproduced in any
form or by any means without permission in writing from the pub-
lisher.
Like having ice cream after a tonsillectomy, this section is my treat after
the book's completion.
Some of those who deserve recognition for their help are addressed
in Chapter 1 in my discussion of my candlestick education. There are
many others whom I would like to thank for their help along my candle-
stick path. Candles might help light the way, but without the assistance
and insights of many others it would have been almost impossible to do
this book. There were so many who contributed in one way or another
to this project that if I have forgotten to mention anyone I apologize for
this oversight.
The Market Technicians Association (MTA) deserves special mention.
It was at the library that I first discovered candlestick material
written in English. This material, albeit scant, was extremely difficult to
obtain, but the marvelously complete MTA library had it. This informa-
tion provided the scaffolding for the rest of my candlestick endeavors.
Besides the two English references on candlesticks I mention in
Chapter 1, I also obtained a wealth of information from books published
in Japanese. I would like to thank the following Japanese publishers and
authors for these books that I used as references:
Then there's the team at Merrill Lynch who were so helpful in look-
ing over the manuscript, making suggestions, and providing ideas. John
Gambino, one of the best colleagues anyone can work with provided all
the Elliott Wave counts in this book. Chris Stewart, Manager of Futures
Research, not only read the entire manuscript but provided valuable
suggestions and finely dissected the many, many charts I used. I also
want to thank Jack Kavanagh in compliance who also read the manu-
script. Yuko Song provided extra insights by conveying some of my can-
dlestick questions to her Japanese customers who use candlesticks.
I have included hundreds of charts in this book from various services.
Before I thank all the services that have generously provided use of their
candlestick charts, I want to give plaudits to Bloomberg L.P. and CQG
(Commodity Quote Graphics).
Bloomberg L.P. was among the first on-line services to provide can-
dlestick charts on the American markets. It's too bad I didn't discover
this earlier. I was drawing candlestick charts on my own for years before
I found out about Bloomberg. CQG, an on-line futures charting service,
was also among the first to see the potential of candlestick charts. Within
a few weeks of my first candlestick article, they sent me an alpha test
(this is a high-tech term for the very early stages of software prototype
testing) of their candlestick software for my CQG System One T". Once I
had this software, my candlestick research progressed exponentially.
Most of the charts in this book are courtesy of CQG.
Besides Bloomberg L.P. and CQG, other services that were kind
enough to provide charts are:
T
Commodity Trend Service Charts (North Palm Beach, FL), CompuTrac "
(New Orleans, LA), Ensign Software (Idaho Falls, ID), Future-
T
Source " (Lombard, Ill), and Quick 10-E Financial Information System
(New York, N.Y.).
I want to thank those who took time from their busy schedules to
review the introductions for Part Two of the book. These are: Dan
Acknowledgements
Gramza for the chapter on Market Profile@;Jeff Korzenik for the chapters
on options and hedging; John Murphy for the chapter on volume and
open interest; once again, John Gambino for the chapter on Elliott Wave;
Charles for the chapter on oscillators; Gerard Sanfilippo and
Judy Ganes for the chapter on hedging; and Bruce Kamich for the
English language glossary.
The Nippon Technical Analysts Association (NTAA) deserves utmost
praise for their assistance. Mr. Kojiiro Watanabe at the Tokyo Investment
Information Center helped me to contact NTAA members who have
been especially helpful. They are: Mr. Minoru Eda, Manager, Quantative
Research, Kokusai Securities Co.; Mr. Yasushi Senior Foreign
Exchange Trader at Sumitomo Life Insurance; and Mr. Nori
Senior Analyst, Fidelity Management and Research (Far East). When I
asked them questions via fax I expected just brief answers. But these
three NTAA members took their valuable time to write pages of explana-
tions, complete with drawings. They were wonderful about sharing their
candlestick experiences and insights with me. I also want to thank them
for reading over and providing information for Chapter 2 on the history
of Japanese technical analysis. If there are any mistakes that remain, they
are those that I failed to correct.
I want to thank again "idea a day" Bruce Kamich. Bruce is a friend
and a fellow futures technician. Throughout our 15-year friendship he
has provided me with many valuable ideas and suggestions. Probably
two of the most important were his suggestion that I join the MTA and
his constant haranguing until I agreed to write a book about candle-
sticks.
Then there's the publishing staff of the New York Institute of
Finance. They were all great, but those with whom I worked most
closely deserve extra praise. Susan Barry and Sheck Cho patiently, skill-
fully and affably guided a neophyte author through the labyrinth of the
book publishing business.
Of course there is my family. At the time that I was writing this book,
our newborn son Evan entered the picture (with all the excitement about
candlesticks, I came close to calling him Candlesticks Nison). Try writing a
book with a newborn and a rambunctious four-year-old daughter, Rebec-
ca, and you start to get an idea of how much my wife, Bonnie, contributed
to this book. She cared for the children while I maladroitly pummeled away
at the keyboard. Obviously, she had the harder job.
For each chapter's heading, and throughout the book, I used Japa-
nese proverbs or sayings. Many times proverbs in the United States are
considered trite and are rarely used. This is not so in Japan where prov-
erbs are respected. Besides being enjoyable to read, the Japanese prov-
erbs offer insights into Japanese beliefs and perspectives. I would like to
Acknowledgements
thank the following publishers for the use of their material for the prov-
erbs and sayings used in this book: University of Oklahoma Press,
Charles E. Tuttle, and Kenkyusha Ltd.
Finally, I must give proper and legal acknowledgements to many of
the services I relied upon during my writing and research. Tick Volume
T
Profile " is a registered trademark of CQG. Market and Liquid-
ity Data Bank@are registered trademarks of the Chicago Board of Trade.
The CBOT holds exclusive copyrights to the Market ProfileB and Liquid-
ity Data Bank@graphics. Graphics reproduced herein under the permis-
sion of the Chicago Board of Trade. The views expressed in this
publication are solely those of the author and are not to be construed as
the views of the Chicago Board of Trade nor is the Chicago Board of
Trade in any way responsible for the contents thereof.
PREFACE
Preface ix
Chapter INTRODUCTION 1
Some background, 1
How I learned about candlestick charts, 1
Why have candlestick charting techniques captured
the attention of traders and investors around the world?, 4
What is in this book, 5
Some limitations, 7
The importance of technical analysis, 8
Chapter STARS 55
The morning star, 56
The evening star, 59
The morning and evening doji stars, 64
The shooting star and the inverted hammer, 70
The inverted hammer, 75
CONCLUSION
BIBLIOGRAPHY
INDEX
Contents
INTRODUCTION
SOME BACKGROUND
"Why," I have often asked myself, "has a system which has been
around so long almost completely unknown in the West?" Were the Jap-
anese trying to keep it secret? Was it the lack of information in the
United States? I don't know the answer, but it has taken years of
research to fit all the pieces together. I was fortunate in several ways.
Introduction
By the time you read this book, there probably will be additional services
providing candlestick charts. Their popularity grows stronger every day.
The profusion of services offering the candlestick charts attests to both
their popularity and their usefulness.
I have had calls and faxs from around the world requesting more infor-
mation about candlestick techniques. Why the extensive interest? There
are many reasons and a few are:
1. Candlestick charts are flexible. Users run the spectrum from first-time
chartists to seasoned professionals. This is because candlestick charts
can be used alone or in combination with other technical analysis
techniques. A significant advantage attributed to candlestick charting
techniques is that these techniques can be used in addition to, not
instead of, other technical tools. I am not trying to convince veteran
technicians that this system is superior to whatever else they may be
using. That is not my claim. My claim is that candlestick charting
techniques provide an extra dimension of analysis.
2. Candlestick charting techniques are for the most part unused in the
United States. Yet, this technical approach enjoys a centuries-old tra-
dition in the Far East, a tradition which has evolved from centuries of
trial and error.
3. Then there are the picturesque terms used to describe the patterns.
Would the expression "hanging-man line" spark your interest? This is
only one example of how Japanese terminology gives candlesticks a
flavor all their own and, once you get a taste, you will not be able to
do without them.
4. The Japanese probably know all the Western methods of technical
analysis, yet we know almost nothing about theirs. Now it is our turn
to benefit from their knowledge. The Japanese use a combination of
candlestick charting techniques along with Western technical tools.
Why shouldn't we do the same?
5. The primary reason for the widespread attention aroused by candle-
stick charts is that using them instead of, or in addition to, bar charts
is a win-win situation.
Introduction 5
Part I of the book reveals the basics on constructing, reading, and inter-
preting over 50 candlestick chart lines and patterns. Part explains how
to meld candlestick charts with Western technical analysis techniques.
This is where the true power of candlecharts is manifested. This is how
I use them.
I have drawn illustrations of candlestick patterns to assist in the edu-
cational process. These illustrations are representative examples only.
The drawn exhibits should be viewed in the context that they show cer-
tain guidelines and principles. The actual patterns do not have to look
exactly as they do in the exhibits in order to provide the reader with a
valid signal. This is emphasized throughout the book in the many chart
examples. You will see how variations of the patterns can still provide
'mportant clues about the state of the markets.
Thus, there is some subjectivity in deciding whether a certain candle-
stick formation meets the guidelines for that particular formation, but
this subjectivity is no different than that used with other charting tech-
niques. For instance, is a $400 support area in gold considered broken if
prices go under $400 intra-day, or do prices have to close under
Does a penetration of $400 substantiate broken support or is a larger
penetration needed? You will have to decide these answers based on
your trading temperment, your risk adversity, and your market philoso-
phy. Likewise, through text, illustrations and real examples I will pro-
vide the general principles and guidelines for recognizing the candlestick
formations. But you should not expect the real-world examples to always
match their ideal formations.
6 Introduction
the next bull move, I equate "surpassed" with "on a close above." That
is because, to me, a close is more important than an intra-day move
above a candlestick line. Another example of subjectivity: In the Japa-
nese literature many candlestick patterns are described as important at a
high-price area or at a low-price area. Obviously what constitutes a
"high-price" or "low-price" area is open to interpretation.
SOME LIMITATIONS
In order to drive home the point about the importance of mass psy-
chology, think about what happens when you exchange a piece of paper
called "money" for some item like food or clothing? Why is that paper,
with no intrinsic value, exchanged for something tangible? It is because
of a shared psychology. Everyone believes it will be accepted, so it is.
Once this shared psychology evaporates, when people stop believing in
money, it becomes worthless.
Second, technicals are also an important component of disciplined
trading. Discipline helps mitigate the nemesis of all traders, namely,
emotion. As soon as you have money in the market, emotionalism is in
the driver's seat and rationale and objectivity are merely passengers. If
you doubt this, try paper trading. Then try trading with your own
funds. You will soon discover how deeply the counterproductive aspects
of tension, anticipation, and anxiety alter the way you trade and view
10 Introduction
may be discounted when the event occurs. Thus, current prices should
reflect all available information, whether known by the general public or
by a select few.
NOTES
'Hill, Julie Skur. "That's Not What I Said," Business Tokyo, August 1990, pp.
'Smith, Adam. The Money Game, New York, NY: Random House, 1986, p. 154.
Bob. The New Gatsbys, Chicago, Bob Tamarkin, 1985, pp. 122-123.
CHAPTER
HISTORICAL
BACKGROUND
social rank. The Bakufu was apprehensive about the increasing amount
of power acquired by certain merchants. In 1642, certain officials and
merchants tried to corner the rice market. The punishment was severe:
their children were executed, the merchants were exiled, and their
wealth was confiscated.
The rice market that originally developed in Yodoya's yard was insti-
tutionalized when the Dojima Rice Exchange was set up in the late 1600s
in Osaka. The merchants at the Exchange graded the rice and bargained
to set its price. Up until 1710, the Exchange dealt in actual rice. After
1710, the Rice Exchange began to issue and accept rice warehouse
receipts. These warehouse receipts were called rice coupons. These rice
receipts became the first futures contracts ever traded.
Rice brokerage became the foundation of Osaka's prosperity. There
were more than 1,300 rice dealers. Since there was no currency standard
(the prior attempts at hard currency failed due to the debasing of the
coins), rice became the defacto medium of exchange. A daimyo needing
money would send his surplus rice to Osaka where it would be placed
in a warehouse in his name. He would be given a coupon as a receipt
for this rice. He could sell this rice coupon whenever he pleased. Given
the financial problems of many daimyos, they would also often sell rice
coupons against their next rice tax delivery (taxes to the daimyo were
paid in rice-usually 40% to 60% of the rice farmer's crop). Sometimes
the rice crop of several years hence was mortgaged.
These rice coupons were actively traded. The rice coupons sold
against future rice deliveries became the world's first futures contracts.
The Dojima Rice Exchange, where these coupons traded, became the
world's first futures exchange. Rice coupons were also called "empty
rice" coupons (that is, rice that was not in physical possession). To give
you an idea of the popularity of rice futures trading, consider this: In
1749, there were a total of 110,000 bales (rice used to trade in bales) of
empty-rice coupons traded in Osaka. Yet, throughout all of Japan there
were only 000 bales of
Into this background steps Homma, called "god of the markets."
"Munehisa Homma was born in 1724 into a wealthy family. The Homma
family was considered so wealthy that there was a saying at that time,
"I will never become a Homma, but I would settle to be a local lord."
When Homma was given control of his family business in 1750, he began
trading at his local rice exchange in the port city of Sakata. Sakata was a
collections and distribution area for rice. Since Homma came from
Sakata, you will frequently come across the expression "Sakata's Rules"
in Japanese candlestick literature. These refer to Homma.
16 A Historical Background
NOTES
first name is sometimes translated as Sokyu and his last name is sometimes translated as
Honma. This gives you an idea of the difficulty of translating Japanese into English. The same
Japanese symbols for Homma's first name, depending on the translator, can be Sokyu or
hisa. His last name, again depending on the translator, can be either Homma or Honma. I chose
the English translation of Homma's name as used by the Nippon Technical Analysts Association.
Johannes and Yui, Tsunehiko. The Development of JapaneseBusiness 1600-1973, Cam-
bridge, MA: Harvard University Press, 1975, p. 31.
PART
THE BASICS
CONSTRUCTING
THE CANDLESTICKS
Since candlestick charts are new to most Western technicians, the most
common Western chart, the bar chart, is used throughout this chapter as
an instructional tool for learning how to draw the candlestick lines.
Drawing the daily bar chart line requires open, high, low, and close.
The vertical line on a bar chart depicts the high and low of the session.
The horizontal line to the left of the vertical line is the opening price. The
horizontal line to the right of the vertical line is the close.
Exhibit 3.3 shows how the same data would be used to construct a
bar chart and a candlestick chart. Although the daily bar chart lines and
candlestick chart lines use the same data, it is easy to see that they are
drawn differently. The thick part of the candlestick line is called the real
The Basics
. .
. .
......................................
. . . I t . .
...
. .
'Jan 'Feb
122
Candlestick Chart
1 2 3 4 5 1 2 3 4 5
Time Period Time Period
body. It represents the range between that session's opening and closing.
When the real body is black filled in) it means the close of the ses-
sion was lower than the open. If the real body is white empty), it
means the close was higher than the open.
The thin lines above and below the real body are the shadows. These
shadows represent the session's price extremes. The shadow above the
real body is called the upper shadow and the shadow under the real body
is known as the lower shadow. Accordingly, the peak of the upper shadow
is the high of the session and the bottom of the lower shadow is the low
of the session. It is easy to see why these are named candlestick charts
since the individual lines often look like candles and their wicks. If a
candlestick line has no upper shadow it is said to have a shaven head. A
candlestick line with no lower shadow has a shaven bottom. To the Japa-
nese, the real body is the essential price movement. The shadows are
usually considered as extraneous price fluctuations.
Exhibits 3.4 through 3.7 demonstrate some common candlestick lines.
Exhibit 3.4 reveals a long black candlestick reflecting a bearish period in
which the market opened near its high and closed near its low. Exhibit
3.5 shows the opposite of a long black body and, thus, represents a bull-
ish period. Prices had a wide range and the market opened near the low
and closed near the high of the session. Exhibit 3.6 shows candlesticks
having small real bodies and, as such, they represent a tug of war
between the bulls and the bears. They are called spinning tops and are
neutral in lateral trading bands. As shown later in this book (in the
24 The Basics
High
Close
Open
LOW'
L o w,------
points of the trading day. The Japanese have a proverb that says, "the
first hour of the morning is the rudder of the day." So is the opening the
rudder for the trading session. It furnishes the first clue about that day's
direction. It is a time when all the news and rumors from overnight are
filtered and then joined into one point in time.
The more anxious the trader, the earlier he wants to trade. Therefore,
on the open, shorts may be scrambling for cover, potential longs may
want to emphatically buy, hedgers may need to take a new or get out of
an old position, and so forth.
After the flurry of activity on the open, potential buyers and sellers
have a benchmark from which they can expect buying and selling. There
are frequent analogies to trading the market and fighting a battle. In this
sense, the open provides an early view of the battlefield and a provi-
sional indication of friendly and opposing troops. At times, large traders
may try to move the market on the open by executing a large buy or sell
order. Japanese call this a morning attack. Notice that this is another mil-
itary analogy. The Japanese use many such military comparisons as we
shall see throughout the book.
The other pivotal price point is the close. Margin calls in the futures
markets are based on the close. We can thus expect heavy emotional
involvement into how the market closes. The close is also a pivotal price
point for many technicians. They may wait for a close to confirm a break-
out from a significant chart point. Many computer trading systems (for
example, moving average systems) are based on closes. If a large buy or
sell order is pushed into the market at, or near, the close, with the
intention of affecting the close, the Japanese call this action a night attack.
Exhibits 3.4 to 3.7 illuminate how the relationship between a period's
open, high, low, and close alters the look of the individual candlestick
line. Now let us turn our attention to how the candlestick lines, alone or
in combination, provide clues about market direction.
CHAPTER
REVERSAL PATTERNS
Technicians watch for price clues that can alert them to a shift in mar-
ket psychology and trend. Reversal patterns are these technical clues.
Western reversal indicators include double tops and bottoms, reversal
days, head and shoulders, and island tops and bottoms.
Yet the term "reversal pattern" is somewhat of a misnomer. Hearing
that term may lead you to think of an old trend ending abruptly and
then reversing to a new trend. This rarely happens. Trend reversals usu-
ally occur slowly, in stages, as the underlying psychology shifts gears.
A trend reversal signal implies that the prior trend is likely to change,
but not necessarily reverse. This is very important to understand. Com-
pare an to a car traveling forward at 30 The car's red
brake lights go on and the car stops. The brake light was the reversal
indicator showing that the prior trend (that is, the car moving forward)
was about to end. But now that the car is stationary will the driver then
decide to put the car in reverse? Will he remained stopped? Will he
decide to go forward again? Without more clues we do not know.
Exhibits 4.1 through 4.3 are some examples of what can happen after
a top reversal signal appears. The prior for instance, could con-
vert into a period of sideways price action. Then a new and opposite
trend lower could start. (See Exhibit 4.1.) Exhibit 4.2 shows how an old
can resume. Exhibit 4.3 illustrates how an can abruptly
reverse into a downtrend.
It is prudent to think of reversal patterns as trend change patterns. I
was tempted to use the term "trend change patterns" instead of "rever-
sal patterns" in this book. However, to keep consistent with other
28 The Basics
EXHIBIT 4.1. Top Reversal EXHIBIT 4.2. Top Reversal EXHIBIT 4.3. Top Reversal
Exhibit 4.4 shows candlesticks with long lower shadows and small real
bodies. The real bodies are near top of the daily range. The variety
of candlestick lines shown in the exhibit are fascinating in that either line
can be bullish or bearish depending on where they appear in a trend. If
either of these lines emerges during a downtrend it is a signal that the
downtrend should end. In such a scenario, this line is labeled a hammer,
I I
White or I
, Reversal Patterns
or Black
29
1. The real body is at the upper end of the trading range. The color of
the real body is not important.
2. A long lower shadow should be twice the height of the real body
3. It should have no, or a very short, upper shadow.
The longer the lower shadow, the shorter the upper shadow and the
smaller the real body the more meaningful the bullish hammer or bear-
ish hanging man. Although the real body of the hammer or hanging
man can be white or black, it is slightly more bullish if the real body of
the hammer is white, and slightly more bearish if the real body of the
hanging man is black. If a hammer has a white real body it means the
market sold off sharply during the session and then bounced back to
close at, or near, the session's high. This could have bullish ramifica-
tions. If a hanging man has a black real body, it shows that the close
could not get back to the opening price level. This could have potentially
bearish implications.
It is especially important that you wait for bearish confirmation with
30 The Basics
the hanging man. The logic for this has to do with how the hanging-man
line is generated. Usually in this kind of scenario the market is full of
bullish energy. Then the hanging man appears. On the hanging-man
day, the market opens at or near the highs, then sharply sells off, and
then rallies to close at or near the highs. This might not be the type of
price action that would let you think the hanging man could be a top
reversal. But this type of price action now shows once the market starts
to sell off, it has become vulnerable to a fast break.
If the market opens lower the next day, those who bought on the open
or close of the hanging-man day are now left "hanging" with a losing
position. Thus, the general principle for the hanging man; the greater the
down gap between the real body of the hanging-man day and the open-
ing the next day the more likely the hanging man will be a top. Another
bearish verification could be a black real body session with a lower close
than the hanging-man sessions close.
Exhibit 4.7 is an excellent example of how the same line can be bear-
ish (as in the hanging-man line on July 3) or bullish (the hammer on July
Although both the hanging man and hammer in this example have
black bodies, the color of the real body is not of major importance.
Exhibit 4.8 shows another case of the dual nature of these lines.
There is a bearish hanging man in mid-April that signaled the end of the
. . .
-Hanging
Man
240
230
Hammer
. . .
220
EXHIBIT 4.7. Soybean Oil-December, 1990, Daily (Hanging Man and Hammer)
Reversal Patterns 31
.
Hanging
Hanging
Man
. . .
................ 26000
. . .
...
,
'Feb 'Mar
122
EXHIBIT 4.8. Dow Jones Industrials-1990, Daily (Hanging Man and Hammer)
rally which had started with the bullish hammer on April 2. A variation
of a hanging man emerged in mid-March. Its lower shadow was long,
but not twice the height of the real body. Yet the other criteria (a real
body at the upper end of the daily range and almost no upper shadow)
were met. It was also confirmed by a lower close the next day. This line,
although not an ideal hanging man, did signal the end of the upturn
which started a month earlier. Candlestick charting techniques, like
other charting or pattern recognition techniques, have guidelines. But,
they are not rigid rules.
As discussed above, there are certain aspects that increase the impor-
tance of hanging-man and hammer lines. But, as shown in the hanging
man of mid-March, a long lower shadow may not have to be twice the
height of the real body in order to give a reversal signal. The longer the
lower shadow, the more perfect the pattern.
Exhibit 4.9 shows a series of bullish hammers numbered 1 to 4 (ham-
mer 2 is considered a hammer in spite of its minute upper shadow). The
interesting feature of this chart is the buy signal given early in 1990. New
lows appeared at hammers 3 and 4 as prices moved under the July lows
at hammer 2. Yet, there was no continuation to the downside. The bears
had their chance to run with the ball. They fumbled. The two bullish
32 The Basics
hammers (3 and 4) show the bulls regained control. Hammer 3 was not
an ideal hammer since the lower shadow was not twice the height of the
real body. This line did reflect, however, the failure of the bears to
maintain new lows. The following week's hammer reinforced the conclu-
sion that a bottom reversal was likely to occur.
In Exhibit 4.10 hammers 1 and 3 are bottoms. Hammer 2 signaled the
end of the prior downtrend as the trend shifted from down to neutral.
Hammer 4 did not work. This hammer line brings out an important
point about hammers (or any of the other patterns I discuss). They
should be viewed in the context of the prior price action. In this context,
look at hammer 4. The day before this hammer, the market formed an
extremely bearish candlestick line. It was a long, black day with a shaven
head and a shaven bottom (that is, it opened on its high and closed on
its low). This manifested strong downside momentum. Hammer 4 also
punctured the old support level of January 24. Considering the afore-
mentioned bearish factors, it would be prudent to wait for confirmation
that the bulls were in charge again before acting on hammer 4. For
example, a white candlestick which closed higher than the close of ham-
mer 4 might have been viewed as a confirmation.
Drawing the intra-day chart using candlesticks shows the high, low,
open, and close of the session (see Exhibit 4.11). For example, an hourly
- -
Reversal Patterns
. . .
Hammer 4
. . .
. . . . .
. .
:
. . .
'Feb 'Mar
122
1800
1890 . Hammer .
1895
9 4/16
session would have a candlestick line that uses the opening and close for
that hour in order to determine the real body. The high and low for that
hour would be used for the upper and lower shadows. By looking
closely at this chart, one can see that a hammer formed during the first
hour on April 11. Like hammer 4 in Exhibit 4.10, prices gapped lower but
the white candlestick which followed closed higher. This helped to con-
firm a bottom.
The second hourly line on April 12, although in the shape of a ham-
mer, was not a true hammer. A hammer is a bottom reversal pattern.
One of the criterion for a hammer is that there should be a downtrend
(even a minor one) in order for the hammer to reverse that trend. This
line is not a hanging man either since a hanging man should appear after
an In this case, if this line arose near the highs of the prior
black candlestick session, it would have been considered a hanging man.
Exhibit 4.12 shows a hammer in early April that successfully called
the end of the major decline which had began months earlier. The long
lower shadow, (many times the height of the real body) a small real
body, and no upper shadow made this a classic hammer.
Exhibit 4.13 shows a classic hanging-man pattern. New highs were
made for the move via an opening gap on the hanging-man day. The
. . . . . . 'Jul .
11612313016
market then gaps lower leaving all those new longs, who bought on the
hanging man's open or close, left "hanging" with a losing position.
In Exhibit 4.14 we see that the rally, which began in early February,
terminated with the arrival of two consecutive hanging-man lines. The
importance of bearish confirmation after the hanging-man line is
reflected in this chart. One method of bearish confirmation would be for
the next day's open to be under the hanging man's real body. Note that
after the appearance of the first hanging man, the market opened higher.
However, after the second hanging man, when the market opened
under the hanging man's real body, the market backed off.
Exhibit 4.15 illustrates that a black real body day, with a lower close
after a hanging-man day, can be another method of bearish confirma-
tion. Lines 1, 2, and 3 were a series of hanging-man lines. Lack of bear-
ish confirmation after lines 1 and 2 meant the was still in force.
36 The Basics
1016 . . . . . . .
1017
1003
1005 'Feb
Hanging
Man
A= \ 'Hanging
20001
30 27 25
Source: L.P.
a very small real body, no upper shadow, and a long lower shadow. The
next day's black real body confirmed this hanging man and indicated a
time to vacate longs. (Note the bullish hammer in early April.)
ENGULFING PATTERN
The hammer and hanging man are individual candlestick lines. As pre-
viously discussed, they can send important signals about the market's
health. Most candlestick signals, however, are based on combinations of
individual candlestick lines. The engulfing pattern is the first of these
multiple candlestick line patterns. The engulfing pattern is a major rever-
sal signal with two opposite color real bodies composing this pattern.
Exhibit 4.18 shows a bullish engulfing pattern. The market is in a
trend, then a white bullish real body wraps around, or engulfs, the prior
period's black real body. This shows buying pressure has overwhelmed
selling pressure. Exhibit 4.19 illustrates a bearish engulfing pattern. Here
the market is trending higher. The white real body engulfed by a black
body is the signal for a top reversal. This shows the bears have taken
over from the bulls.
There are three criteria for an engulfing pattern:
EXHIBIT 4.18. Bullish Engulfing Pattern EXHIBIT 4.19. Bearish Engulfing Pattern
1. If the first day of the engulfing pattern has a very small real body and
the second day has a very long real body. This would reflect a dissi-
pation of the prior trend's force and then an increase in force behind
the new move.
2. If the engulfing pattern appears after a protracted or very fast move.
A protracted trend increases the chance that potential buyers are
already long. In this instance, there may be less of a supply of new
longs in order to keep the market moving up. A fast move makes the
market overextended and vulnerable to profit taking.
3. If there is heavy volume on the second real body of the engulfing
pattern. This could be a blow off (volume using candlestick charts is
discussed in Chapter 15).
40 The Basics
4. If the second day of the engulfing pattern engulfs more than one real
body.
Exhibit 4.20 shows that the weeks of May 15 and May 22 formed a
bullish engulfing pattern. During the last two weeks of July, a bearish
engulfing pattern emerged. September's bullish engulfing pattern was
the bottom of the prior to the major rally.
In Exhibit 4.21 a monthly crude oil chart with both the bullish and
bearish engulfing patterns can be seen. In late 1985, a precipitous $20
decline began. The third and fourth month of 1986 showed the two can-
dlestick lines of the bullish engulfing pattern. It signaled an end to this
downtrend. The rally that began with this bullish engulfing pattern con-
cluded with the bearish engulfing pattern in mid-1987. The small bullish
engulfing pattern in February and March of 1988 terminated the
trend that started with the mid-1987 bearish engulfing pattern. After this
bullish engulfing pattern, the trend went from down to sideways for five
months.
The black candlestick of February 1990 came within 8 ticks of engulf-
ing the January 1990 white candlestick. Consequently, this was not a
perfect bearish engulfing pattern but, with candlesticks, as with other
Reversal Patterns 41
. . . . . . . . . . . . . . . . . . . . . . .
Bearish Engulfing
Patterns
.
Source:
, .................. .................... ......,............... ......... ... . ... .... . ...................:........... ..... .........:...
,,
EXHIBIT 4.23.
Soybeans-Weekly
I Jul Jan Jul (Bearish Engulfing
Patterns)
Source: Copyright 1990 Commodity Trend Servicee
DARK-CLOUD COVER
Our next reversal pattern is the dark-cloud cover (see Exhibit 4.24). It is a
two candlestick pattern that is a top reversal after a or, at times,
at the top of a congestion band. The first day of this two candlestick pat-
tern is a strong white real body. The second day's price opens above the
The Basics
prior session's high (that is, above the top of the upper shadow). How-
ever, by the end of the second day's session, the market closes near the
low of the day and well within the prior day's white body. The greater
the degree of penetration into the white real body the more likely a top
will occur. Some Japanese technicians require more than a 50% penetra-
tion of the black session's close into the white real body. If the black
candlestick does not close below the halfway point of the white candle-
stick it may be best to wait for more bearish confirmation following the
dark cloud cover.
The rationale behind this bearish pattern is readily explained. The
market is in an A strong white candlestick is followed by a gap
higher on the next session's opening. Thus far, the bulls are in complete
control. But then no continuation of the rally occurs! In fact, the market
closes at or near the lows of the day moving well within the prior day's
real body. In such a scenario, the longs will have second thoughts about
their position. Those who were waiting for selling short now have a
benchmark to place a stop-at the new high of the second day of the
dark-cloud cover pattern.
The following is a list of some factors that intensify the importance of
dark-cloud covers:
1. The greater the degree of penetration of the black real body's close
into the prior white real body, the greater the chance for a top. If the
black real body covers the prior day's entire white body, a bearish
engulfing pattern would occur. The dark-cloud cover's black real
body only gets partially into the white body. Think of the dark-cloud
cover as a partial solar eclipse blocking out part of the sun (that is,
covers only part of the prior white body). The bearish engulfing pat-
tern can be viewed as a total solar eclipse blocking out the entire sun
(that is, covers the entire white body). A bearish engulfing pattern,
consequently, is a more meaningful top reversal. If a long, white real
body closes above the highs of the dark-cloud cover, or the bearish
engulfing pattern, it could presage another rally.
Reversal Patterns
. . . . . . ..
Dark-cloud
Cover 3
4/90
1855
.:. ...................................................................................
:. ........
1873 Feb 'Mar 'May
I
white real body. The municipal bond market backed off after this top
reversal appeared. The final coup de grace came a few weeks later when
the bearish engulfing pattern materialized. We see how the dark-cloud
cover's black real body covered only part of the prior white real body.
The black real body of the bearish engulfing pattern enveloped the entire
previous white real body.
In Exhibit 4.26 three dark-cloud covers can be seen. Other bearish
signals confirmed each of these patterns. Let us look at them on an indi-
vidual basis.
15 SPMO D A I L Y BAR
Cover 2 Dark-cloud ,
. . . Dark-cloud .
Cover 1
. .
I . . . . .
Cover
.......................
dark-cloud cover is the result of a new high on the open, with the mar-
ket closing deeply into the prior white real body. What would happen,
though, if, on the second day of the dark-cloud cover, the open pene-
trates the highs not from days, or even weeks ago, but from months ago
and then fails at these new highs? This would produce very negative
connotations. This is the scenario that unfolded in April. The highest
levels in at least three months were touched on the black candlestick
session of dark-cloud cover 3. This high failed to hold and prices closed
well within the prior white real body.
In Exhibit 4.28, we see that the price incline commencing February 10
came to an abrupt halt with the mid-February dark-cloud cover.
PIERCING PATTERN
push higher, creating a relatively long, white real body that closes above
the mid-point of the prior day's black real body.
The bullish piercing pattern is akin to the bullish engulfing pattern.
In the bullish engulfing pattern the white real body engulfs the previous
black real body. With the bullish piercing pattern, the white real body
only pierces the prior black body. In the piercing pattern, the greater the
degree of penetration into the black real body, the more likely it will be
a bottom reversal. An ideal piercing pattern will have a white real body
that pushes more than halfway into the prior session's black real body.
If the market closes under the lows of the bullish engulfing pattern or
the piercing pattern by way of a long black candlestick, then another
should resume.
The psychology behind the piercing pattern is as follows: The market
is in a downtrend. The bearish black real body reinforces this view. The
next day the market opens lower via a gap. The bears are watching the
market with contentment. Then the market surges toward the close,
managing not only to close unchanged from the prior day's close, but
sharply above that level. The bears will be second guessing their posi-
tion. Those who are looking to buy would say new lows could not hold
and perhaps it is time to step in from the long side.
The piercing pattern signal increases in importance based on the
same factors (1) through (4) as with the dark-cloud cover, but in reverse.
(See previous section.) In the section on the dark-cloud cover, I men-
tioned that although some Japanese traders like to see the black real
body close more that midway in the prior white candlestick, there is
some flexibility to this rule. With the piercing pattern, there is less flexi-
bility. The piercing pattern's white candlestick should push more than
halfway into the black candlestick's real body. The reason for less lati-
tude with the bullish piercing pattern than with the bearish dark-cloud
cover pattern is the fact that the Japanese have three other patterns
called the on-neck, the in-neck, and the thrusting pattern (see Exhibits 4.30
to 4.32) that have the same basic formation as the piercing pattern, but
which are viewed as bearish signals since the white real body gets less
than halfway into the black's real body.
50 The Basics
Piercing
I I
3700
3600
. . .
3500
Piercing , ,
4/26/90
. Pattern
. Variation Pattern
.
.
is, a shaven bottom) and closed its high (that is, a shaven head). Note
how this bullish piercing pattern brought to an end the that com-
menced with the bearish engulfing pattern of March 19 and 20.
On this Wheat chart there is also a variation of the piercing pattern
during the week of March 12. The reason it is a variation is because the
white real body opened under the prior day's real body, but not under
the prior day's low. Nonetheless, because the white real body closed
more than 50% into the prior day's black real body it was a warning sign
that the prior was running out of steam.
Exhibit 4.35 illustrates how candlestick patterns can help the analyst
get a quick sense of the market's health. During the latter part of Febru-
ary 1990, a broker asked me what I thought of oats. I rarely monitor oats.
Nonetheless, I retrieved the candlestick chart shown in Exhibit 4.35 and
told him that the downtrend was probably over. Why? I had noticed that
during the week of February 20, an almost classic piercing pattern
appeared. I also saw this piercing pattern coincided with a successful test
of the early February lows. This increased the chance that a double bot-
tom had been built.
Exhibit 4.36 illustrates that the downtrend, which began with the
bearish engulfing pattern in late 1984, ended in mid-1987 with the
44 D A I L Y BAR 1989
..
,
. . . . . .
. . .
. . .
. . . . . .
. . .
. . .
. . .
. . .
. . .
1551
. .. .
Pattern
'Feb 'Mar
appearance of this piercing pattern. Although the market did not rally
after this bottom reversal signal, the signal did forecast the end of the
selling pressure that had pulled the market down from mid-1984 to mid-
1987. After the piercing pattern the market stabilized for a year, and then
rallied.
CHAPTER
STARS
a
"One cannot be too cautious"
Star
(can be
white or
black)
(can be
I
I
In In Downtrend In In Downtrend
In any of these star patterns the real body of the star can be white or
black.
The morning star (see Exhibit 5.3) is a bottom reversal pattern. Its name
is derived because, like the morning star (the planet Mercury) that fore-
tells the sunrise, it presages higher prices. It is comprised of a tall, black
real body followed by a small real body which gaps lower (these two
lines comprise a basic star pattern). The third day is a white real body
that moves well within the first period's black real body. This pattern is
a signal that the bulls have seized control. I will break down this three-
candlestick pattern into its components in order to understand the ratio-
nale behind this last statement.
.. ..
.. ..
.. . .
.. ..
. 7500
.. ..
Dark-cloud .
. ' . Cover
..
. . . ..
. . . . .
. . .
. . .
..
.. ..
1/11/90 , . . . .. Star ..
, 6590
6690 .. ..
6590
6665 'Nov 'Jan
(27 (11
Oct Jan
EXHIBIT 5.5. Crude Oil-Weekly
(Morning Star) Source: @Copyright 1990 Trend Servicea
Stars 59
EXHIBIT 5.6.
September 1990, Daily
(Morning Star)
I
I I
EXHIBIT 5.7. Evening Star
60 Basics
...............
............................
..................
:..
.
..
: . :
,
I . i..
Jan
1988
II""
Jul
to yellow (the star's warning signal) to red (the black real body confirms
the prior trend has stopped).
In principle, an evening star should have a gap between the first and
second real bodies and then another gap between the second and third
real bodies.' However, from my experience this second gap is rarely seen
and is not necessary for the success of this pattern. The main concern
should be the extent of the intrusion of the third day's black real body
into the first day's white real body.
At first glance Exhibit 5.7 is like an island top reversal as used by
Western technicians. Analyzing the evening star more closely shows it
furnishes a reversal signal not available with an island top (see Exhibit
5.8). For an island top, the low of session 2 has to be above the highs of
sessions 1 and 3. However, the evening star only requires the low of the
real body 2 to be above the high of real body 1 to be a reversal signal.
The evening star pattern shown in Exhibit 5.9 reflects the Summer of
1987 which called the high of the Dow just before the crash. (I wonder if
the Japanese technicians who use candlesticks were looking at this!)
Exhibit 5.10 provides an example of how candlestick indicators can
transmit a reversal signal not easily found with Western tools. The last
............................
9/12
6311 .
6316 .
6298 .
, 6305 .
--
I 4 9/10
hour on September 5 and the first two hours the next day formed an
evening star pattern. The star portion of this evening star pattern would
not have been an island top based on the aforementioned discussion. In
this instance, candlesticks provided a top reversal indication not avail-
able with the Western island top. Also note how the rally that ended
with this evening star began with the morning star on September 4.
Although more important after an the evening star can be
important at the top of a congestion band if it confirms another bearish
signal. (See Exhibit 5.11.) That is what happened in the middle of April.
The star portion (that is, the second day) of the evening star coincided
with a resistance area. The basis for this resistance at $413 was that it
was an old support level from late March. Old support often converts to
new resistance. Try to remember this! It is a very useful trading rule.
Chapter 11 discusses support and resistance in more detail. In any case,
Mar
Source:
the resistance level near $413 coincided with the appearance of the
evening star thus reinforcing the negativeness of the pattern.
Exhibit 5.12 shows a well-defined evening star in mid-December. The
star was preceded by a strong, white real body and followed by a weak,
black real body. A variation of an evening star appeared in mid-
November. The reason it was a variation is that the evening star usually
has a long, white real body preceding the star, and then a black real
body after the star. We did not see the long, white or black real body
lines here. We view this as a top, however, not only because of its minor
resemblance to an evening star pattern, but because of the hanging-man
line on November 21 (the "star" portion of the evening star). The next
day's opening under the hanging man's real body confirmed a top.
Some factors that would increase the likelihood that an evening or
morning star could be a reversal would include:
1. If there is a gap between the first candlestick's and star's real bodies
and then in the star's and third candlestick's real bodies;
2. If the third candlestick closes deeply into the first candlestick's real
body;
64 The Basics
................................................................................
: A NOTE
The full name of the evening and morning star patterns are the three-river
evening star and the three-river morning star. I originally thought they were
termed "three-river" evening and morning stars because each of these
: patterns had three candlestick lines-hence three rivers. I discovered that
the origin is much more fascinating.
Nobunaga Oda, a major military figure of the late 16th century, was
: one of the three military leaders who unified feudal Japan (see Chapter 2).
He fought a seminal battle that occurred in a very fertile rice growing
province. Since rice was a gauge of wealth, Nobunaga was as determined
to wrest this area as fervently as the owners were to defend it. This fertile
rice area had three rivers. The heavily defended area made it difficult for
: Nobunaga to cross these three rivers. Victory was his when his forces
finally forded these three rivers. Hence the name "three river" morning :
and evening star where it is difficult to change the trend. Yet, victory for
: the attacking army is assured when the hurdle of the "three rivers" is
crossed.
................................................................................
3. If there is light volume on the first candlestick session and heavy vol-
ume on the third candlestick session. This would show a reduction of
the force for the prior trend and an increase in the direction force of
the new trend.
When a doji gaps above a real body in a rising market, or gaps under a
real body in a falling market, that doji is called a doji star. Exhibit 5.2
shows doji stars. Doji stars are a potent warning that the prior trend is
apt to change. The session after the doji should confirm the trend rever-
sal. Accordingly, a doji star in an followed by a long, black real
body that closed well into the white real body would confirm a top
reversal. Such a pattern is called an evening doji star (see Exhibit 5.13).
The evening doji star is a distinctive form of the regular evening star.
The regular evening star pattern has a small real body as its star (that is,
the second candlestick), but the evening doji star has a doji as its star.
The evening doji star is more important because it contains a doji.
A doji star during an is often the sign of an impending top.
It is important to note that if the session after the doji star is a white can-
dlestick which gaps higher, the bearish nature of the doji star is negated.
Stars 65
EXHIBIT 5.13. Evening Doji Star EXHIBIT 5.14. Morning Doji Star
Abandoned Baby
435
.......
..............................
I
I
. . 400
Star
.......................................................
I
. 395
EXHIBIT 5.17.
I I
December 1989, Daily 23
(Morning Doji Star) Source: Bloomberg L.P.
black real body before and the white real body after the doji star made
this three-line pattern a morning doji star.
On the doji star candlestick of Exhibit 5.18, prices broke under
This was a support area from early in July. The fact that the new lows
could not hold is considered bullish. Add to this the morning doji star
pattern and you have two reasons to suspect a bottom.
Exhibit 5.19 is an example of both an evening doji star and a regular
evening star. Price action from March through May 1986 formed an
evening doji star. This pattern halted a sharp rally which began just a
few months previously. A ensued after this evening doji star. It
ended with the bullish engulfing pattern. The rally from that engulfing
pattern topped during the evening star pattern of mid-1987.
In Exhibit 5.20, we see the three lines that form the evening doji star
on March 17, 18, and 19. This pattern ended the rally that began with a
hammer the prior week. This example again shows that certain candle-
stick configurations should have more latitude in the equity market. This
is because, unlike futures, stock prices may open relatively unchanged
from the prior close. This means that specific patterns that relate the
open to the prior day's close may have to be adjusted for this fact.
In the case of Dow Chemical, note how the evening doji star was not
a true star. A doji star's real body (that is, its opening and closing price)
should be over the prior day's real body. Here it was not. Therefore,
allow more flexibility with candlestick indicators with equities. For those
who monitor the equity markets, as you experiment with candlestick
techniques, you should discover which patterns may have to be modi-
fied.
In Exhibit 5.21, one can see that a few weeks before 1987's major
off, an evening doji star top arose. The center candlestick of this pattern
(the doji star) did not gap above the prior white candlestick as should a
true star. However, as discussed in Exhibit 5.20, one should allow more
latitude with this concept of gaps since stocks often open at, or very
near, the prior session's close.
Exhibit 5.22 reveals a very unusual and ominous occurrence in that
back-to-back evening doji patterns formed. Candlestick lines through 3
formed an evening doji star. The next three sessions, lines 4 through 6,
fashioned another evening doji star.
...
...
.
...
...
,
....
... 90
.
.: .
I" "
Jul Jan Jan EXHIBIT 5.21.
1987 1988 NYSE-Weekly
Source: 1990 Commodity Trend Servicee (Evening Doji Star).
. , . , ,
520C
..
...........
. . . . .
8/31/90
4925 .
4930 .
4800
4868
120
White or Black
6/14/90 . ..
..
'Nay 'Jun
Shooting Star 1
...............
18
Source: Bloomberg L.P.
Exhibit 5.26 is another example where the shooting star pattern did
not gap away from the prior real body. It was, nonetheless, a significant
reversal signal. Here again let us look at the shooting star in context. It
was another failure at the third quarter 1989 highs. The shooting star
spelled the end of a rally that began with the hammer.
Exhibit 5.27 reveals that a classic shooting star made its appearance
in the first hour of May 29. The ensuing price decline stopped at the
bullish engulfing pattern on June 4.
Exhibit 5.28 illustrates two shooting stars that preceded meaningful
price declines. Exhibit 5.29 shows that the shooting star was also a fail-
ure at the 1989 highs. A double whammy! Exhibit
5.30 shows a pair of shooting stars. Each spelled the end of the preced-
ing rally.
Stars 73
........... \............'.......-
I I I I
JUN 15 29 27
6/90
29486
28272
, 28599 Ju Oct Ju
INVERTED HAMMER
While not a star pattern, we'll discuss the inverted hammer in this sec-
tion because of its resemblance to the shooting star. Exhibit 5.31 illus-
trates that an inverted hammer looks like a shooting star line with its long
upper shadow and small real body at the lower end of the range. But,
while the shooting star is a top reversal line, the inverted hammer is a
bottom reversal line. As with a regular hammer, the inverted hammer is
a bullish pattern after a downtrend.
Refer back to the corn chart discussed in Exhibit 5.27. Look at the first
candlestick of the bullish engulfing pattern of June 4. It has the same
appearance as the shooting star (the color of the real body does not
76 The Basics
2600
..
250C
Piercing .
.....
lnverted
Hammer .
4/90
2472 .. . .
2485
2472
2485 'Nay 'Jun
EXHIBIT 5.32. London Zinc-Three Month, 1990 (Shooting Stars and Inverted Hammer)
:
lnverted
. . Inverted Hammer
Hammer .. 2
1
..
, .. ..
. 8020 ,
The pattern (see Exhibit 6.1) is a small real body which is con-
tained within a prior relatively long real body. "Harami" is an old Japa-
nese word for "pregnant." The long candlestick is "the mother"
candlestick and the small candlestick as the "baby" or "fetus." In Chap-
ter 3, we discussed how spinning tops (that is, small real bodies) are
useful in certain formations. The harami is one of these formations (the
star, examined in Chapter 5, is another).
The Basics
White
I In In Downtrend
small real body is inside the larger one. The size of the shadows are usu-
ally not important in either a harami or harami cross.
The harami displays a disparity about the market's health. After a
bull move, the long white real body's vitality is followed by the small
real body's uncertainty. This shows the bulls' upward drive has weak-
ened. Thus a trend reversal is possible. During a bear move, the heavy
selling pressure reflected by a long, black real body is followed by the
second day's vacillation. This could portend a trend reversal since the
second day's small real body is an alert that the bears' power has dimin-
ished.
Exhibit 6.4 illustrates that a small rally started on April 18. Harami 1
........................ ..........
4770 Harami 2 .
, 4778
4710
4760 'May 'Jun
116
82 The Basics
Source:
called its end and the that started with harami 1 stopped with
harami 2. Harami 3 reflects how a harami pattern might be useful even
if there is no evident trend before a harami pattern occurs. Note that
there was no evident trend during the first few days of May. Then
harami 3 arose with its long, white real body followed by a small, black
real body (remember the color of the second day's real body is not
important).
A trader could, nonetheless, use this pattern as a signal that the rally
More Reversal Formations 83
started on the strong, white day had failed. The market was now at a
point of indecision. A buy would not be recommended until the indeci-
sion had been resolved via a close above the highs of harami 3.
Harami 4 was a classic. An was evident prior to the tall white
candlestick. The next day's small real body completed the harami. This
small real body also took on the negative aspects of a shooting star day
(although not a perfect star since the real body was not above the prior
real body).
Exhibit 6.5 illustrates exemplary harami. Each of the second day's real
bodies are diminutive compared to the prior long real bodies. The first
harami implied a lack of upside momentum; the second harami implied
a drying up of selling pressure.
Exhibit 6.6 illustrates how the two candlestick harami pattern in late
March spelled the top of the market. The continued until the bull-
ish hammer occurred on April 24. Notice how the shadow of the second
session in the harami was outside the real body of the prior session. This
demonstrates the importance of the relationship between the real bodies
and not the shadows.
Exhibit 6.7 shows a steep decline which ensued from the bearish
engulfing pattern of May 7 and 8. This harami marked the change of a
downtrend into a lateral band.
Intra-day traders could use the harami in Exhibit 6.8 as a signal that
the prior intra-day trend might be over. Appropriate action would then
be warranted. In this example, the early April 17 precipitous price
decline ended and the market went into a lull after the harami pattern.
This harami could have been used by day traders to cover shorts. Like
30
Source: Bloomberg L.P.
9300
.................................................................................................................................
4/17
any bottom reversal pattern, this harami did not preclude the possibility
that the market would resume its downward course. Yet, this harami
relayed a condition about the market. Specifically, it told us that, at least
at the time of the harami, the downward pressure had subsided.
Exhibit 6.9 is a good example of a precipitous downtrend converted
to a lateral trading environment after the advent of the harami. In this
example we see how the prior downtrend, in which prices cascaded
from $5.40 to $4.85, stopped at the harami. But the harami did not nec-
essarily imply a rally. After a harami the market usually eases into a
congestion band.
Harami Cross
The regular harami has a tall real body followed by a smaller real body.
Yet, there are no rules as to what is considered a "small" candlestick.
This, like many other charting techniques, is subjective. As a general
principle, the smaller the second real body, the more potent the pattern.
This is usually true because the smaller the real body, the greater the
ambivalence and the more likely a trend reversal. In the extreme, as the
86 The Basics
. .
6600
real body becomes increasingly smaller as the spread between the open
and close narrows, a doji is formed.
As mentioned, a doji preceded by a long real body is called a harami
cross. The harami cross carries more significance than a regular harami
pattern. Where the harami is not a major reversal pattern, the harami
cross is a major reversal pattern. A harami cross occurring after a very
long white candlestick is a pattern a long trader ignores at his own peril.
Harami crosses also call bottoms, but they are more effective at tops.
Exhibit 6.10 illustrates how the rally from mid-March abruptly ended
when the harami cross pattern formed on April 2 and 3. Exhibit 6.11
shows how the large upside gap made in mid-January shouted, "bull
market." But, the harami cross said, "no bull market now." Exhibit 6.12
shows how an unusually large black candlestick session followed by a
doji created a harami It shows how the market had severed itself
from the prior downtrend. A hammerlike session after the doji of the
harami cross (that successfully tested the recent lows) gave further proof
of a bottom.
More Reversal Formations 87
:. I.
......................................
.. . . .
I
.. .. . . .
2/15/90
2494 .. ..
, 2506
2486
2502 'Jan 'Feb
. 7350
EXHIBIT 6.12. Live
Cattle-April 1990,
Source: Ensign Software Daily (Harami Cross)
88 The Basics
Tweezers are two or more candlestick lines with matching highs or lows.
They are called tweezers because they are compared to the two prongs
of a tweezers. In a rising market, a tweezers top is formed when the highs
match. In a falling market, a tweezers bottom is made when the lows are
the same. The tweezers could be composed of real bodies, shadows,
doji. A tweezers occurs on nearby or consecutive sessions and as
such are usually not a vital reversal signal. They take on extra impor-
tance when they occur after an extended move or contain other bearish
(for a top reversal) or a bullish (for a bottom reversal) candlestick signals.
Exhibits 6.13 through 6.18 elaborate on this idea.
Exhibit 6.13 shows how, in an a long white line is followed
by a doji. This two-candlestick pattern, a harami cross with the same
high, can be a significant reversal signal. Exhibit 6.14 illustrates a twee-
zers top formed by a long white candlestick and a hanging-man line
during the next session. If the market opens under the hanging-man's
real body, odds are strong that a top has been reached. The market
should not close above the tweezers top in order for this bearish view to
prevail. This two-line mixture can also be considered a harami. As such,
it would be a top reversal pattern during an Exhibit 6.15 illus-
trates a tweezers top joined with the second period's bearish
star line. Although not a true shooting star, the second line is bearish
based on the price action which creates it; the market opens near its
lows, rallies to the prior session's high, then closes near its low for the
session. This would also be considered a harami pattern.
Exhibit 6.16 illustrates a variation on the dark-cloud cover. Here, the
second day opens above the prior day's close (instead of above the prior
day's high). The black candlestick day's high touches the prior period's
high and then falls. This could be viewed as a combination of a
cloud cover and a tweezers top. Exhibit 6.17 shows a hammer session
which successfully tests the prior long black lows. The
hammer, and the successful test of support, proves that the sellers are
losing control of the market. This two-line combination can also be
viewed as a harami. This would be another reason to view this action as
important support. Exhibit 6.18 is a variation on the bullish piercing line
with a tweezers bottom added in for good measure. A true piercing line
would open under the prior day's low. Here it opened under the prior
day's close.
These examples of tweezers are not inclusive. They are representative
of how top and bottom tweezers should be confirmed by other candle-
stick indications so as to be valuable forecasting tools. For those who
want a longer time perspective, tweezers tops and bottoms on the
weekly and monthly candlestick charts made by consecutive candlesticks
could be important reversal patterns. This would be true even without
other candlestick confirmation because, on a weekly or monthly chart,
for example, a low made the preceding session that is successfully tested
this session could be an important base for a rally. Less important, and
less likely to be a base for a rally, would be, on a daily chart, a low made
yesterday that is successfully tested today.
Exhibit 6.19 illustrates tweezer tops and bottoms. The tweezers top
was confirmed when the second day completed a bearish engulfing pat-
tern. Tweezers bottom pattern 1 illustrates a star. Note also how this
two-day tweezer bottom was a successful test of the piercing pattern
from the prior week. Tweezers bottom 2 is a set of two hammers. The
combination of these two bullish indications, the tweezers bottom and
the hammers, set the stage for a rally.
Exhibit 6.20 shows that the lows made on January 24, near were
retested a week later. The test was not only successful, but this test built
a bullish engulfing pattern. Exhibit 6.21 shows that on February 14 and
15, a two-day tweezers bottom also established a bullish engulfing pat-
tern. Exhibit 6.22 illustrates a hanging man following a long white can-
dlestick. The highs on both of these weeks (as well as the next) were the
same, thus creating a tweezers top. The two lines of the tweezers were
also a harami pattern. Exhibit 6.23 shows that a variation of an evening
star developed in late June. For a true evening star, we like to see a
longer white candlestick as the first line in the pattern. Nonetheless, this
90 The Basics
CPKO D R I L Y BAR
. . . .
.. . . .
1
Tweezers
.....................................
Hararni
Hanging
Man
Tweezers Too
Evening
Star
.........
16 30
Source: Bloomberg L.P.
... . and Do
BELT-HOLD LINES
The belt hold is an individual candlestick line which can be either bull-
ish or bearish. The bullish belt hold is a strong white candlestick which
opens on the low of the day (or with a very small lower shadow) and
moves higher for the rest of the day. The bullish belt-hold line is also
called a white opening shaven bottom. If, as in Exhibit 6.25, the market is at
a low price area and a long bullish belt hold appears, it forecasts a rally.
The bearish belt hold (see Exhibit 6.26) is a long black candlestick which
opens on the high of the session (or within a few ticks of the high) and
continues lower through the session. If prices are high, the appearance
of a bearish belt hold is a top reversal. The bearish belt-hold line is
sometimes called a black opening shaven head.
More Reversal Formations
I on
High
Opens on
I
Low
EXHIBIT 6.25. Bullish Belt Hold EXHIBIT 6.26. Bearish Belt Hold
The longer the height of the belt-hold candlestick line, the more sig-
nificant it becomes. Belt-hold lines are also more important if they have
not appeared for a while. The actual Japanese name for the belt hold is
a sumo wrestling term: yorikiri. It means "pushing your opponent out of
the ring while holding onto his belt." A close above a black bearish
hold line should mean a resumption of the A close under the
white bullish belt-hold line implies a renewal of selling pressure.
Exhibit 6.27 shows how bullish belt-hold line 1 signaled a rally.
hold line 2 is interesting. It confirmed a tweezers bottom since it main-
tained the prior week's lows. A rally ensued which ended with a harami
a few weeks later.
........... ..........
Hold 1
. Bullish Belt .
, 7/23/90 Hold 2
4830
4830
4759
4775- 'Ju 'Oc t 'Jan 'Jul
.. ..
7/26/70 ,
7320 .. ..
7360 ,
7275
7345 Jun
The shooting star was the first sign of trouble in Exhibit 6.28. The
next session's bearish belt-hold line confirmed a top. Another bearish
belt hold during the following week reflected the underlying weakness
of the market.
Exhibit 6.29 is an example of back-to-back bearish belt holds in mid-
February. The which ensued, was sharp, but brief as a bullish
morning star pattern spelled a bottom.
More Reversal Formations
EXHIBIT 6.30. Upside Gap Two Crows EXHIBIT 6.31. Mat-hold Pattern
More Reversal Formations
Source:
dlestick's upper shadow or closes above the last black candlestick's high,
then buying is warranted. This pattern can have two, three, or four black
candlesticks. The upside gap two crows and the mat-hold are relatively
rare.
Exhibit 6.32 is a good example of this type of upside-gap two crows
pattern. In early February, the two crows flew above a long white can-
dlestick. This pattern called an end to the rally which had begun a
month earlier.
In Exhibit 6.33, one can see that on November 27, copper pushed
ahead via a long white session. New highs on the two following sessions
failed to hold. The second black candlestick session made this into an
upside-gap two crows pattern. The market slid until a doji star and a
tweezers bottom built a platform for another leg higher.
Exhibit 6.34 is a classic example of the rare mat-hold pattern. A strong
white candle followed by a black candlestick that gaped higher. Two
more small black candlesticks followed with the white candlestick com-
pleting the mat-hold pattern. Note how this pattern is not too much dif-
ferent from the upside-gap two crows (remember the mat hold can also
have two, instead of three, small black candlesticks just as the
The Basics
I
...................................................................................
Upside-gap
T w o Crows
. .
.
.
.
. .
.
. . .
. . .
. . .
..
..
. \I
.. Tweezer . ..
. Bottom
.. ..
.. ..
.. ..
: 950C
.. ..
10900 'Jan 'Feb
(20 1 (22 (29
... .
I""
.
Jan
1987
gap two crows has). The main difference is the appearance of the white
candlestick at the end which turns the pattern bullish. Thus, for an
upside-gap two crows, I suggest you place a stop on a close above the
second black candlestick's high.
The upside-gap two crows consists of two black candlesticks. If there are
three declining consecutive black candlesticks it is called three black crows
pattern (see Exhibit 6.35). The three black crows presage lower prices if
they appear at high-price levels or after a mature advance. Three crows
are also sometimes called three-winged crows. The Japanese have an
expression, "bad news has wings." This is an appropriate saying for the
three-winged crow pattern. The three crows are, as the name implies,
three black candlesticks. Likened to the image of a group of crows sitting
ominously in a tall dead tree, the three crows have bearish implications.
The three lines should close at, or near, their lows. Each of the openings
should also be within the prior session's real body. The analyst would
also like to see the real body of the first candlestick of the three crows
under the prior white session's high.
Exhibit 6.36 is a good example of a three crows pattern. In mid-June,
three black crows appeared. Another three crow pattern shows up a
month later in mid July. July's three crows also was a failure at the highs
from June's three crows near the 33,000 level. This formed a double top.
Exhibit 6.37 is another example of this pattern. June 15 was the first of
the three crows. An interesting aspect about these three crows is that the
open of the second and third black candlesticks are at, or very near, the
close of the prior black candlestick. This is referred to as identical three
crows. It is regarded as especially bearish, but it is a very rare pattern.
I
EXHIBIT 6.35. Three Crows
The Basics
Three . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .. . .
: :..................... ........
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
....................................................................................................................
. . .
. . . :. . . . : . . .
'Jun 'Jul 'Nov
122
................................ 9200
..
: ..........
. . ..
7/26/90 .
8945
8973
8930
114 121
COUNTERATTACK LINES
I
6.39. Bearish Counterattack
to the prior day's close. The bulls' tide of optimism on the second day's
opening probably turned to apprehension by the close.
As the bullish counterattack line is related to the piercing line, so the
bearish counterattack line is related to the dark-cloud cover. The bearish
counter attack, like the dark-cloud cover, opens above the prior day's
high. Unlike the dark-cloud cover, though, the close does not go into the
prior day's white candlestick. Thus, the dark-cloud cover sends a stron-
ger top reversal signal than does the bearish counterattack line.
An important consideration of these counterattack lines is if the sec-
ond session should open robustly higher (in the case of the bearish
counterattack) or sharply lower (for the bullish counterattack). The idea
is that on the opening of the second day of this pattern, the market has
moved strongly in the direction of the original trend. Then, surprise! By
the close, it moves back to unchanged from the prior 'session!
On May 29, in Exhibit 6.40, the long white candlestick reinforced the
bullish outlook from a rally that started the prior week. Sure enough, on
May 30, the market surged ahead on the opening. However, it was
downhill from there for the rest of the session. By the close, the market
had fallen back to unchanged from the prior close. These two sessions,
May 29 and 30, constructed a bearish counterattack pattern.
Exhibit 6.41 shows that a rally terminated with the bearish counterat-
tack line. Exhibit 6.42 shows that the price waterfall, which began with
the bearish engulfing line in March 1989, ended with the bullish coun-
terattack a few months later. Remember, all trend reversal indicators,
like the counterattack line, tells you is that the trend will change. That
does not mean prices will reverse direction. Here is an example of how
a bullish reversal pattern signaled that the prior downtrend was over as
the trend went from down to sideways. This example also shows that
the closes do not have to be identical in order to make the pattern valid.
In Exhibit 6.43 one can see how prices eroded from the shooting star
until the bullish counterattack line appeared. Another positive feature
about this bullish counterattack line is that it was a session which
More Reversal Formations 105
.. ..
.. ..
.. Bearish
Bearish
.. , Counterattack ,
. Line
..
. .
. .
. . ..
. . ..
. .Mau .
Bearish
Counterattack
Line
. Pattern
Bullish
Counterattack
4/90 Line
5050
5095
5040
5070 'Jul 'Oc 'Jan
opened under the late July, early August support area but, nonetheless,
the new lows could not be maintained. This showed that the bears could
not take control of the market.
EXHIBIT 6.46. Three River Bottom EXHIBIT 6.47. Inverted Three Buddha
book Technical Analysis of Stock Trends, much of the material in that book
is based on Schabacker's work. Schabacker was Edward's father-in-law.)
It is intriguing how market observers from both the West and the East
have come up with the this same pattern. Market psychology is the same
the world 'round, or, as a Japanese proverb expresses, "The tone of a
song is the same everywhere."
The three river bottom pattern (see Exhibit 6.46) is the opposite of the
three mountain top. This occurs when the market tests a bottom level
three times. The peak of the troughs should be exceeded to confirm a
bottom. The equivalent of the Western head and shoulders bottom (also
called an inverted head and shoulders) is called the modified three river bottom
pattern or the inverted three Buddha pattern (see Exhibit 6.47).
Exhibit 6.48 is an unusual chart in that it has the various manifesta-
tions of the three mountain top. They are as follows:
man followed by two doji, and area 3 is another bearish engulfing pat-
tern. Since the central mountain formed the highest peak in Exhibit 6.50,
this pattern became a three Buddha pattern. The black real body within
the prior white real body formed a harami at the peak of the high central
mountain.
As illustrated in Exhibit 6.51, there was an inverted three Buddha
pattern in 1988 (that is, similar to an inverted head and shoulders). Each
of these bottoms at A, B, and C had a bullish candlestick indication. At
A, a hammer appeared. At B, another hammer appeared that was part
of a morning star pattern (the morning star rally ended with the
cloud cover). At C, a piercing line appeared (it was almost a bullish
engulfing pattern). Once the bulls gapped above the downward sloping
resistance line, the trend turned up. Gaps are called windows by Japanese
technicians. (Windows will be discussed at length in the next chapter on
continuation patterns.) Because I refer to gaps in this chapter, the reader
should note that the Japanese view gaps (that is, windows) as continua-
tion patterns. Thus a gap higher is bullish and a gap lower is bearish. In
this example, the gap higher had bullish implications. The price activity
from the third quarter of 1989 into the first quarter of 1990 effected a
three Buddha top.
In order for the three river bottoms (including this one) to provide a
buy signal, there should be a close via a white candlestick above the
peaks of the troughs (see Exhibit 6.52). In this case, it would be above
More Reversal Formations
B
Inverted Three Buddha,
2013
.Jan .Jan
I1990
,
EXHIBIT 6.51. Crude Oil-Weekly (Inverted Three Buddha Bottom and Three Buddha Top)
The emphasis on triple tops and bottoms by the Japanese probably has
to do with the importance of the number three in the Japanese culture.
We, as Westerners, would not necessarily see anything special about the
three peaks. Our view would be that double tops and, more rarely, tops
which have been tested four times can be just as significant as triple
tops. But the Japanese think differently. And maybe they can show us a
side of Western analysis which we may have overlooked. Intriguingly,
there are many pattern and technical concepts based on the number
three in Western technical analysis as well as in candlestick charting. The
following is a quote from John Murphy's book Technical Analysis of the
Futures Markets:
It's interesting to note how often the number three shows up in the study
of technical analysis and the important role it plays in so many technical
approaches. For example, the fan principle uses three lines; major bull
and bear markets have three phases (Dow theory and Elliott Wave The-
ory); there are three kinds of gaps some of the more commonly known
reversal patterns, such as the triple top and the head and shoulders, have
three prominent peaks; there are three different classifications of trend
(major, secondary, and minor) and three trend directions (up, down, and
sideways); among the generally accepted continuation patterns, there are
three types of triangles-the symmetrical, ascending, and descending;
there are three principle sources of information-price, volume and open
interest. For whatever the reason, the number three plays a very promi-
nent role throughout the entire field of technical analysis.'
three black crows that portend a price fall; top patterns include the
three mountain top and its variation; the three Buddha pattern; the three
river bottoms; the three windows (see Chapter 7) which define the extent
of a move; the three methods (see Chapter 7); and the three candlestick
patterns including the morning and evening stars. The Japanese also
believe that if a window (during a rising market) is not closed within
three days the market will rally.
The dumpling top (see Exhibit 6.53) usually has small real bodies as the
market forms a convex pattern. When the market gaps down, confirma-
tion of a dumpling top occurs. This pattern is the same as the Western
rounded bottom top. The dumpling top should have a downside win-
dow as proof of a top.
The fry pan bottom (see Exhibit 6.54) reflects a market which is bottom-
ing and whose price action forms a concave design and then a window
to the upside opens. It has the same appearance as a Western rounded
bottom, but the Japanese fry pan bottom should have a window in an
in order to confirm the bottom.
The rounding top and the small real bodies as the market tops out is
indicative of dumpling top as seen in Exhibit 6.55. Note how the doji
was at the peak of the market with the downside window helping to
confirm the dumpling top pattern. The fact the black candlestick after the
window was a black belt-hold line was another reason for a bearish out-
look. Exhibit 6.56 shows a fry pan bottom whose low points on April 27
and 28 formed a harami pattern. A window in early May substantiated
that a fry pan bottom had been put into place.
Exhibit 6.57 illustrates a nicely shaped fry pan bottom. The bullish
confirmation came at candlestick 2. Although the market did not form a
window between candlestick 1 and 2, the fact that the high for
stick 1 was $1,000 and the low for candlestick 2 was $997 meant it only
16 27 25
Source: Bloomberg L.P.
The tower top is a top reversal pattern. It occurs while the market is in an
and then a strong white candlestick (or a series of white candle-
sticks) appears. The market's rise then slows and the highs start falling.
The tower top is completed with the appearance of one or more large
black candlesticks (see Exhibit 6.58). This pattern's long candlesticks
resemble tall towers-hence the name.
The tower bottom develops at low price levels. After one or more long
black candlesticks there is a short-term lull. Then one or more large
white candlesticks emerge. This creates a bottom with towers on both
sides (see Exhibit 6.59). That is, long candlesticks on the way down and
long candlesticks on the way up.
Exhibit 6.60 shows that a group of strong white candlesticks
appeared from first quarter until the second quarter of 1987. Then a
series of long black candlesticks surfaced. The tall white candles formed
the left tower while the long black ones the right tower. The three black
candlesticks were also three black crows.
Exhibit 6.61 illustrates fwo patterns-the tower bottom and a rare
bottom reversal which has not been discussed called the unique three river
. . . . . . . . . . . . . . . . . . . . . .
bottom. First, I will review the tower bottom. A long black candlestick on
August 28, some sideways action via narrow candlesticks, and the long
white candlestick September 7 created a tower bottom. The steep price
descent on August 28 erected the left tower and the sharp rally, which
commenced September 7, erected the right tower. Notice the three can-
dlesticks numbered 1, 2, and 3 on August 28 through August 30. These
comprise the extremely rare three river bottom (see Exhibit 6.62).
Its closest relative in candlesticks is the evening star pattern. The unique
three river bottom is a bottom reversal pattern. Its first line is an
118 The Basics
extended black candlestick, the second line is a black real body candle-
stick that closes higher than the first candlestick's close, and the third
line is a very small real body white candlestick. This last line displays a
market whose selling pressure has dried up.
The closest analogy in Western technical terms to the tower reversals
would be the spike, or V, reversals. In the spike reversal, the market is
in a strong trend and then abruptly reverses to a new trend. The tower
top and bottom are similar to the dumpling top and the fry pan bottom.
The major differences are that the towers have long real bodies before
and after the market turns and the dumpling top and fry pan bottom
have windows. The towers usually also have more volatile candlesticks
than the dumpling tops or fry pan bottoms. Do not worry about whether
a top is a tower top of a dumpling top, or if a bottom is a tower bottom,
or a fry pan. All these patterns are viewed as major reversal patterns.
Note
'Murphy, John J. Technical Analysis of the Futures Markets, New York, NY: New York Institute of
Finance, 1986, p. 79.
CHAPTER
CONTINUATION
PATTERNS
WINDOWS
I
Closed Window
But No Selling
Follow Through
................
Window 3
Window 1
. .
. . .
, 6/90
, Engulfing
Pattern
'Feb 'Mar
10 02 D A I L Y BAR CQG
..........................................................................................................................................
. 1604
, 1589
. . .
Bullish Inverted
Engulfing Hammer
. . .
. . . . .
. 1603
. 1604
. 1589
1601 .Mar
should become support. Add to this the support at the window near this
$1.10 area and you have two reasons to expect $1.10 to provide a solid
floor. Throughout March, this area did indeed provide a firm footing for
the bulls.
The Japanese believe windows from a congestion zone, or from a
new high, deserve special attention. See Exhibit 7.4. The early March
window above was a significant break above a month-long conges-
tion zone. Thus there was dual support in the window near The
first was because of the window, the second was because the old resis-
tance area had become support. Notice the solid support this window
provided for the next few months. April 2 and 3 comprised a harami.
This indicates that the prior trend (in this case, a downtrend) had run
out of steam. A bullish engulfing pattern formed a few days later. On
April 16 an inverted hammer appeared. Each of these patterns appeared
near the window's support at
In March 1988, a bullish engulfing pattern presaged a rally. (See
Exhibit 7.5.) A window opened during the rally. The rally progressed
until the bearish counter attack line. The window held as support for five
weeks but the persistence of selling after the window closed annulled
the
. A=
Counter- attack .
16 16 30
Source: L.P.
Thus far, the focus of this section has been on the use of the window
as support or resistance and as a continuation indication. There is
another use. (See Exhibit 7.6.) A window, especially if it is made with a
small black candlestick from a low price congestion area, can mean a
meaningful upside breakout. Exhibit 7.6 illustrates this principle.
Throughout February prices were locked in a relatively tight congestion
band. Between February 24 and 25, a small upside window opened via
a diminutive black candlestick. This window was confirmed as support
the next session. On that session (February the market not only held
the window as a support but produced the strongest type of candlestick
line, a long white candlestick that opens on its low (that is, a bullish
belt-hold line) and closes on its high.
A large window appeared in mid-January as indicated by Exhibit 7.7.
From late January to late February, there were numerous return moves
to this window (corrections go back to the window). Each of these rallies
was short circuited when they got near the resistance level created by the
window.
Look at the Dow in Exhibit 7.8. The "Crash of '87" created a window
in the 2,150 to 2,200 zone. Two criteria were needed to tell us the
trend was over. The first was to close the large window. The second was
for a continuation of the buying force once the window closed. These
criteria were met in early 1989.
Exhibit 7.9 is another example of a window creating resistance. The
124 The Basics
continued after this level. This is because in mid-July the market formed
a window (window 2). The top of the window was Until the bulls
were able to prove their vigor by pushing prices above this window,
going long should be viewed as a high-risk strategy in spite of the morn-
ing star. The morning star did act as support on a test of its low a few
days after it formed. Yet, after trying for a week the bulls failed in their
attempt to close window 2. This told us a new rally was not likely. The
moral of this story is that candlesticks, or any other technical tool,
should be considered in the context of the prevailing trend.
In Exhibit 7.10, we see that the market headed south after Septem-
ber's hanging man and the black line which engulfed it. The window in
late September signaled a continuation of the decline. The window was
126 Basics
1/90
Star
: 1000
1045
. 1055 May
: :
9500
..
8995
, 8937
8947- t 'Jan
990
Window
Window 2
' 6942
7108 'Jul 'Oc t 'Jan
EXHIBIT 7.12. Gold-December 1990, Daily (Windows and Eight New Record Highs)
all the rally attempts during the week following its opening. Another
interesting aspect about the September rally that stalled at window 2 is
that the rally, as shown by the numbered days 1 through 8, made eight
new higher highs. Candlestick theory states that after about eight to ten
new highs or lows, without a meaningful correction, the odds are strong
that a significant correction will unfold. Each new high or new low for
the move is called a "new record high" or "new record low" by the Jap-
anese. Thus the Japanese will say there are ten record highs or lows,
meaning there were a series of ten higher highs or lower lows. If there
are, say, eight new highs without a meaningful correction, the Japanese
refer to the market by using the expression "the stomach is 80% full."
What is interesting about this gold chart is that there were eight record
highs. This gave a warning that a top might be near. The fact that after
these eight record highs the market was at a resistance area set by win-
dow 2 was an extra strong signal to be cautious about the long side of
this market.
The enchanted number three makes yet another appearance in
Exhibit 7.13. Traditional Japanese technical analysis posits that after
three up or down windows, the chances are- strong that a top (in the case
of three windows in an or a bottom (in the case of three win-
Continuation Patterns 129
..
: :
EXHIBIT 7.14. Upward Gapping Tasuki EXHIBIT 7.15. Downward Gapping Tasuki
The real bodies of the two candlesticks in the tasuki gap be should be
about the same size. Both types of tasuki gaps are rare.
Look at Exhibit 7.16 for an example of an upside-gap tasuki. In the
last week of September, the market experienced a small upside gap via
a white candlestick. The next weekly black candlestick opened in the
prior week's real body and closed under that white real body's open.
This created an upside-gap tasuki. Note how the small window opened
by this pattern provided support in the October pullback. The bullish
belt hold signaled the advent of the rally.
Continuation Patterns 131
Window
It is normal after a sharp one to two session advance for the market to
consolidate the gains. Sometimes this consolidation is by a series of small
real bodies. A group of small real bodies after a strong session tells us
that the market is undecided. However, if there is an upside gap on the
opening (that is, a window) from these small real bodies, it is time to
buy. This is high-price gapping play pattern (see Exhibit 7.17). It is called
this because prices hover near their recent highs and then gaps to the
upside.
A low-price gapping play is, not surprisingly, the bearish counterpart of
the high-price gapping play. The low-price gapping play (see Exhibit
7.18) is a downside window from a low-price congestion band. This
congestion band (a series of small real bodies) initially stabilized a steep
one to two session decline. At first, this group of small candlesticks gives
the appearance that a base is forming. The break to the downside via a
window dashes these bullish hopes.
Exhibit 7.19 illustrates that in late November, a series of
three small real bodies helped digest the gains of the prior tall white
candlestick session. When sugar gapped up it completed the first
upward-gapping play pattern on this chart. The market rallied until the
dark-cloud cover on November 17 and 18. High-price gapping play pat-
tern 2 had a tall white candlestick session, some small real bodies, and
then a window. This window converted to support level.
The candlesticks in Exhibit 7.20 sent a bullish signal when the win-
dow opened on June 29. This window completed the action needed to
form the upward-gapping play. Prior to this gapping play, there was a
strong a strong white candlestick on June 11. A group of small candle-
sticks followed this line. This had the potential to become a high-price
gapping play. No upward gap, however, meant a no buy signal.
As shown in Exhibit 7.21, on July 20 and 21 the S&P quickly fell 18
points. It then traded sideways at lower price levels for over a week (for
132 The Basics
Source:
a gapping play, the consolidation should not last more than 11 sessions).
A Japanese broker related to me that one of her clients in Japan (a fund
manager who applies candlesticks) obtained a sell signal on August 2
(see arrow at the doji) based on the low-price gapping play pattern.
This shows an aspect about the candlesticks which has been dis-
cussed earlier. The techniques and procedures used to interpret the can-
dlestick patterns are guidelines, not hard and fast rules. Here we have
an example where the ideals of the low-price gapping play were not
exactly meet, yet the Japanese fund manager thought it was close
enough to act on. In principle, for a low-price gapping play to be com-
pleted the market should gap lower. The low on August 1 was at 355.80
and the high on August 2 was 355.90. Thus there was no gap. Yet it was
close enough that the Japanese fund manager got his sell signal on
August 2. Note also the sharp decline preceding the small real bodies did
not close on the low. Yet, because prices stayed within the lows of the
range during the next few sessions, it resembled the low-price gapping
play closely enough to provide the Japanese candlestick practitioner with
Continuation Patterns 133
7/26/90 .. I
6000 ,
6040
5982 . .
6040 Ju
..... 36000
..
Low-price
...........................................................................
Play
..
Window
5/31/90
4941
4941
4895
4915 'Mar 'May
(7
white line pattern since the opens on the white candlesticks were not
identical and the white lines were separated by one day. Nonetheless,
this is likened to the side-by-side white line pattern.
In addition, Exhibit 7.24 shows two side-by-side white line
patterns. This pattern portends bullish implications if it emerges at lower
price levels. The first side-by-side white line pattern had three
opens at about the same price. Then the market staged a brief pullback
on May 8 which marginally broke under the window but sprang right
back from there. The second side-by-side white line pattern gave
another bullish signal. As should be the case with the side-by-side
white lines, they provided a firm footing.
These three methods include the bullish rising three methods and the bear-
ish falling three methods. (Note how the number three again makes an
appearance.) These are both continuation patterns. The benchmarks for
the rising three methods pattern (see Exhibit 7.25) include:
The Basics
...
. Rising
I. Three
4257 . . . . . . . . . . . . . . . . . . . . . . . .
. 4260
, 4248
. 4253
.. .. .. .. .. .. .. .. .. .. .. .. .. ..
black candlestick session, the market should head lower. This pattern
resembles a bear flag or pennant formation.
Exhibit 7.27 shows a classic rising three methods pattern. The market
is in an with three downtrending black small real bodies pre-
ceded by a white candlestick. The black candlesticks essentially held
within the white candlestick's range. The last white candlestick then
closes above the first candlestick's close. A factor that may lend more
significance to this pattern is if volume on the white (black) sessions for
the rising (falling) three method pattern is higher than the small candle-
stick sessions. Here we see the white session days of the rising three
methods pattern had greater volume than on the three small black can-
dlestick sessions.
Exhibit 7.28 is also a rising three methods pattern. When completed,
the bonds pushed until they reached the bearish engulfing pattern.
Although the ideal three methods has three small candlesticks follow-
ing a long white one, Exhibit 7.29 is an example of two small candle-
sticks. The price action in June 1988 built a tall white candlestick. Black
candlesticks that held within this white candlestick's range followed in
July and August. September formed another white candlestick which
made a new high for the move but failed to close above June's close by
138 Basics
only 3 ticks. Normally, we would want to see a higher close. In this case,
since the last white candlestick (September) only missed closing above
June's close by 3 ticks, this pattern should still be considered a rising
three methods pattern with bullish confirmation the next session. A new
high close in October gave this confirmation and secured the bullish out-
look.
Three small candlesticks held within the first candlestick's high and
low range are evident in Exhibit 7.30. These were trailed by another
white candlestick. This last white candlestick had the same close as the
first one, so we need confirmation. When the next hour opened above
the last white candlestick, the bullish confirmation was at hand. Observe
how the top of the rising three methods pattern became a support area
as tested by the first hour on August 1.
Two examples of this bullish continuation pattern appear in Exhibit
Continuation Patterns 139
2/90
11429 Three
. . . . . -
12004 . . . Methods
11168
11243
11987 11988 989
7.31. The first rising three methods pattern, in early July, shows how
there can be two candlesticks instead of three, after the first tall white
real body. Notice how the two black candlesticks held within the first
candlestick's range. Then the last white real body of this pattern opened
above the close of the prior session and made a new high close for the
move. The second illustration of this pattern in Exhibit 7.31 displays how
the color of the real bodies after the first candlestick does not have to be
black. As long as the real bodies hold within the first session's range it
has the potential to be a three methods pattern. Here the potential was
fulfilled as the last long white candlestick closed at a new high.
In March 1989, a window appeared as shown in Exhibit 7.32. Based
on the saying that corrections go to the window, one should expect a
bounce up to the window. From there the preceding downtrend should
resume. After the window, three small real bodies developed. The
assault on the window took place on the first week of April. It failed
from there. Two weeks later, on the third small white candlestick week,
there was another attempt to close the window. This attempt also fal-
tered. The last long black candlestick closed under the first black candle-
stick's close. This action completed the five candlesticks of the falling
three methods.
Exhibit 7.33 is an example with four, instead of three, small real bod-
ies. The key is that the real bodies hold within the first day's range. The
last large black candlestick concluded this pattern. Note how tick vol-
ume'" confirmed the black candlesticks action. That is, tick volumeT"
expanded with the black candlesticks and shrank with the intervening
Continuation Patterns 141
6500
Window
Falling .
7/90
6942
7108 'Jul 'Oc 'Jan
--
..
Falling
Three
.T. 11000
DAILY BAR
. 2021
-H= 2021 . .. ..
1991
-L= 2002 . ..
. A = -20
.. ..
2150
. . .
..
.
7/23/90
2021
. . Hammer . ..
.L=
-C= 2002 . . . . . 3
EXHIBIT 7.36. Three Advancing EXHIBIT 7.37. Advance Block EXHIBIT 7.38. Stalled Pattern
White Soldiers
healthy method for the market to rise (although if the white candlesticks
are very extended, one should be cautious about an overbought market).
If the second and third, or just the third candlestick, show signs of
weakening it is an advance block pattern (see Exhibit 7.37). This means that
the rally is running into trouble and that longs should protect them-
selves. Be especially cautious about this pattern during a mature
Signs of weakening could be progressively smaller white real
bodies or relatively long upper shadows on the latter two white candle-
sticks.
If the last two candlesticks are long white ones that make a new high
followed by a small white candlestick, it is called a stalled pattern (see
Exhibit 7.38). It is also sometimes called a deliberation pattern. After this
formation the strength has been at least temporarily exhausted.
This last small white candlestick can either gap away from the long white
body (in which case it becomes a star) or it can be, as the Japanese
express it, "riding on the shoulder" of the long white real body (that is,
be at the upper end of the prior long white real body). The small real
body discloses a deterioration of the bulls' power. The time of the stalled
pattern is the time for the longs to take profits.
Although the advance block and stalled patterns are not normally top
reversal patterns, they can sometimes precede a meaningful price
decline. The advance block and stalled patterns should be used to liqui-
date or protect longs (for example, sell covered calls) but not to short.
They are generally more consequential at higher price levels.
In Exhibit 7.39, the three white soldiers from a low price level in 1985
projected a rally. There follows two advance block patterns. Advance
block 1 has progressively smaller white real bodies in early 1987 which
did not bode well for higher prices. A shooting star was the last small
white real body of this three candlestick group. The market floundered
for the next few months. There was another price push after these doji
but the next advanced block pattern gave another warning sign.
Advance block 2 formed in mid-1987. The major difference between this
three white candlestick pattern and the three white soldiers in advance
Continuation Patterns 145
EXHIBIT 7.39. New York Composite-Monthly (Three White Soldiers, Advance Block)
block 1 is that the last white candlestick had a longer upper shadow. It
was not very long, but it showed that the market did not have the power
to close near the highs. In other words, the lead (that is, advance) sol-
dier had been "blocked." A hanging-man line appeared next month. The
soldiers then went into retreat.
There was another reason to be suspicious about more upside after
advance block 2. Whereas the three white soldiers in 1985 started from a
low price level, the three white candlesticks of the advance block pattern
arose after the market had already sustained an extended advance.
In early 1989 (see Exhibit stalled pattern 1 temporarily stalled
the prior price incline. Additionally, this pattern came after an extended
series of white candles.
Stalled pattern 2 only stalled the advance for a couple of weeks. The
last small real white body in this deliberation pattern was a hanging
man. Once the market closed above the hanging man's high two weeks
later, the market was not likely to fall. In early July, the bullish three
white soldiers started a meaningful rally. It lasted for seven new highs
(that is, seven record highs). Another set of three white soldiers
appeared in the third quarter 1989. Because each of these three white
candlesticks closed at their high, this pattern had all the earmarks of
146 The Basics
A=-148
Star
27500
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25000
White
Soldiers .
Three
. White
Soldiers
: :
10
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . .
Jan Ju Oct Jan
EXHIBIT 7.40. Dow Jones Industrial Average-Weekly (Stalled Patterns and Three White Soldiers)
implying another strong rally, similar to the one that had started in July.
This was not to be. The-week after the third white candlestick of this
group, a small real body emerged. This formed a harami and told us the
prior had run into trouble. The shooting star a few weeks later
confirmed problems at these highs.
Exhibit 7.41 illustrates three stalled patterns. Pattern 1 also formed a
harami which short circuited the rally. Stalled pattern 2 failed to keep the
rally in check, while stalled pattern 3 contained a shooting star. Stalled
pattern 3 induced a trend change as the market went from up, to side-
ways for a few weeks. Remember that the stalled pattern is not usually
a trend reversal, it often means a time of deliberation before the market
decides its next trend. In the case of stalled pattern 3, a window opened
after the congestion band completing a bullish high-price gapping play.
This indicated a resumption of strength.
SEPARATING LINES
Separating Lines
I
Bullish
I EXHIBIT 7.42. Bullish and Bearish
Separating Lines
148 Continuation Patterns
bullish separating line as shown in Exhibit 7.42. The white line should
also be a bullish belt hold (that is, open on the low of the session). The
opposite would be true with the bearish separating line in exhibit 7.42.
This is viewed as a bearish continuation pattern.
THE MAGIC DOJI
sudden danger"
and
Close
EXHIBIT 8.1. Doji EXHIBIT 8.2. Long-legged Doji EXHIBIT 8.3. Gravestone Doji
(Rickshaw Man)
The perfect doji session has the same opening and closing price, yet
there is some flexibility to this rule. If the opening and closing price are
within a few ticks of each other (for example, a cent in grains or a few
thirty-seconds in bonds, and so on), the line could still be viewed as a
150 The Basics
doji. How do you decide whether a near-doji day (that is, where the
open and close are very close, but not exact) should be considered a doji?
is subjective and there are no rigid rules but one way is to look at
a near-doji day in relation to recent action. If there are a series of very
small real bodies, the near-doji day would not be viewed as significant
since so many other recent periods had small real bodies. One technique
is based on recent market activity. If the market is at an important mar-
ket junction, or is at the mature part of a bull or bear move, or there are
other technical signals sending out an alert, the appearance of a near-doji
is treated as a doji. The philosophy is that a doji can be a significant
warning and that it is better to attend to a false warning than to ignore
a real one. To ignore a doji, with all its inherent implications, could be
dangerous.
The doji is a distinct trend change signal. However, the likelihood of
a reversal increases if subsequent candlesticks confirm the doji's reversal
potential. Doji sessions are important only in markets where there are
not many doji. If there are many doji on a particular chart, one should
not view the emergence of a new doji in that particular market as a
meaningful development. That is why candlestick analysis usually
should not use intra-day charts of less than 30 minutes. Less than 30
minutes and many of the candlestick lines become doji or near doji
(except for the very active markets such as bond and futures).
DOJI AT TOPS
Doji are valued for their ability to call market tops. This is especially true
after a long white candlestick in an (see Exhibit 8.4). The reason
for the doji's negative implications in is because a doji repre-
sents indecision. Indecision, uncertainty, or vacillation by buyers will
not maintain an It takes the conviction of buyers to sustain a
rally. If the market has had an extended rally, or is overbought, and then
a doji surfaces (read "indecision"), it could mean the scaffolding of buy-
ers' support will give way.
can be
away from,
or within,
prior candle
I
Doji after
long white
8.4. Doji Following a Tall White Candlestick candlestick
Continuation Patterns 151
..
........
. ...
....
.
...
...
.......
.......
....
.:.
.......
. .. ,. . . . ,. . . . . , . . . . . .
...............................
23 30 6 13 20 27 4 11 18 25 1 8 15 22 29 5 12 19 26 5 12 19 26 2 9
Oct Nov Jan Feb Mar
A=
Doji 1 Doji 2
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
when doji 8 and 9 formed a double bottom was there a trend reversal
(albeit temporary). Thus there may be less need for confirmation of a top
reversal via a doji than for a bottom reversal.
Exhibit 8.6 illustrates that the rally which began in mid-1987, gave its
first sign of peaking with doji 1. Another warning flag was hoisted with
doji 2, a few months later. The hanging man after doji 2 confirmed the
top. A minor rally attempt ended in late 1989 at doji 3. This exhibit
exemplifies that confirmation after a doji increases success in projecting
a trend reversal. The white candlestick, which appeared a month after
doji 1, did not confirm the top hinted at by doji 1. Bearish confirmation
came only after doji 2. After doji 2, this verification came in the form of
a hanging man and then a long black candlestick. Confirmation of doji 3
as a top came with the next month's long black candlestick session.
The more conservative trading style used, the more important it is to
wait for verification of a trend change. How long should one wait for
corroboration? It is a trade-off between risk and reward. If one bases
one's trading style on waiting for trend reversal corroboration less risk
should be involved but, it also could provide less reward. By the time
the reversal is substantiated, profit potential may be reduced.
Exhibit 8.7 shows three doji, each after an Doji 1 signaled a
The Magic Doji 153
6600
. . .
. . .
Doji 1 ,
..
. .
........................ 6000
..
minor top. Doji 2 did not correctly call a reversal, but it was followed the
next day by an decline. Doji 3 is interesting. It is more important
than the prior two doji since it followed a series of three long white can-
dlesticks and it formed a harami cross. Doji 3 strongly stressed that the
prior might be over. When it appeared, longs should have taken
protective measures (the prior strong negated short selling).
This means that they should either be liquidating some longs, moving
up protective stops levels, selling calls.
An intra-day spike higher the next day made it appear that the pre-
diction about the end of the was going to be wrong. But, on that
day, the market sold off sharply toward the close. This action helped
confirm the original view that the prior was about over. The
market then went into a congestion phase for the next few weeks. A
pattern resembling an evening star then arose. It was not an ideal
evening star pattern since the star portion did not gap away from the
prior long white real body, yet it presaged a top.
154 The Basics
Exhibit 8.8 shows that a doji after a long white candlestick, especially
after a prolonged is often a forewarning that a top is near. This
exhibit has three examples of this concept:
2850
2800
2750
2500
2450
10 17 24 31 7 14 21 28 4 11 18 25 2 9 16 23 30 6 13 20 27 4 11 18 25 1 8 15 22 29 5 12 19
Nov Jan Feb
Source Trend
. : . . .
. . . . : .: :. . . . : . : . .
. . . . . . .
13 46 WEEKLY BAR
9024
9108
, 9006
.
I I
"""I'
, 9305
. 8927
Ju Oct Jan Ju
Doji .
EXHIBIT 8.11. Cotton-July, 1990, Intra-day (Doji after a Long White Candlestick)
The Magic Doji 157
. . . . .
5000
..
..
..
..
4900 ..
4900
4830
4850 'Mar 'Nay
Exhibit 8.13 has a strong hint of a peak with the long-legged doji
(here the opening and closing were close enough to consider this a doji
session). The long-legged doji day also completed a harami pattern and
a tweezers top. This confluence of technical factors were forceful clues
that the highs were at hand. Exhibit 8.14 illustrates that a price peak in
gold was reached with the long-legged doji in January. The long upper
shadows in early February confirmed the resistance set by the long-
legged doji.
The Magic Doji 159
Long Legged
an Mar I
Gravestone ,
. . .
. . .
3/90
9114
9115
9114
9115 'Feb
. . . . Doji
Pattern
..............
that the gravestone doji on October 8 (the very minor lower shadow
does not void this as a gravestone) was especially negative for this stock.
On that day, a new high was touched. It was the bulls' chance to propel
prices, but they failed. By the close, prices had pulled back to near the
daily low. There was trouble at this $41 level before. Beginning on Sep-
tember 29, three candlestick lines developed into a stalled pattern. The
gravestone doji confirmed the heavy supply at
Some of you may have noticed that a gravestone doji looks like a
shooting star. The gravestone doji, at tops, is a specific version of a
shooting star. The shooting star has a small real body, but the grave-
stone doji-being a doji-has no real body. The gravestone doji is more
bearish than a shooting star.
Rickshaw M a n .
: : : : :
of the doji week in September 1989 became a support area. The doji star
top in late September became a resistance level.
In Exhibit 8.18, the rickshaw man (the real body was small enough to
consider it a doji) on the first hour of March 21 gave a clue that the pre-
vious could be reversing. A doji occurring a few hours later give
more proof for this outlook. These two doji became a significant resis-
tance area.
THE TRI-STAR
The tri-star (see Exhibit 8.19) is a very rare, but a very significant rever-
sal pattern. The tri-star is formed by three doji lines. The middle doji is
a doji star. I have yet to see an ideal tri-star, as shown in Exhibit 8.19,
but the following examples show the significance of this pattern even in
its variations. The reason we are discussing this pattern here, instead of
The Magic 163
I
Tri-star Top Tri-star Bottom
EXHIBIT 8.19. Tri-star Top and Bottom
in the chapter on stars, is because the most important aspect of this pat-
tern is that the lines should be three doji (or near-doji).
Exhibit 8.20 shows that early in the week of September 15 there were
two doji followed by a third, small real body candlestick. This variation
on a tri-star was the start of a rally. Exhibit 8.21 shows that in late
September 1989, the Dow began a rally that culminated in a series of
three doji in early October. Although not an ideal tri-star, the three doji
after a 170-point advance was a portentous sign. Notice that the latter
two doji also formed a tweezer top.
.................................
I
.
I I
SEP 15 29
Source: L.P.
SCROLL
Source: Future Source"
PUTTING IT ALL
TOGETHER
Part I of this book, we've examined many candlestick lines and for-
mations. This chapter is a visual summary. The following charts (see
Exhibits 9.1, 9.2, and 9.3) have numbered lines and patterns. All are
candlestick indicators that have been discussed previously. How would
you interpret them? If necessary use the visual Japanese candlestick glos-
sary at the back of the book to help you with your interpretations. My
opinion of these candlestick patterns and lines are provided. But, you
should decide for yourself.
Remember, the following interpretations are subjective. You may see
different indicators than I did, or some where I did not. As with any
charting technique, different experiences will give different perspectives.
There are no concrete rules, just general guidelines. For example, what
if a hammerlike line had a lower shadow only one and one-half times the
height of the real body instead of the more ideal version with a lower
shadow twice-or even three times-the height of the real body. A pur-
est might say this was not a hammer and ignore it. Others may cover
shorts on such a line. Still others might wait for the next session to see
what unfolds.
166 The Basics
9. The three-day rally that started with the inverted hammer in (8) was
halted by this harami pattern.
10. A hammer signaled a possible bottom.
A variation on the bullish piercing pattern occurred. Instead of the
second white candlestick opening under the first day's low it opens
under the first day's close. It then rallies and closes well into the
black candlesticks real body.
12. Another hanging man occurs. But this line was not confirmed by the
next session since the market gapped higher on the open.
13. A bearish engulfing pattern occurred.
14. Then, a classic piercing pattern presented itself. The second session
of this pattern was a bullish belt-hold line that closed on its high. It
also successfully tested the lows made in (11).
15. Then, a doji star signaled an end to the rally which started at (14).
16. A harami called the end of the preceding price descent.
9/90
.. . .. . ... . . . . . . . . .... . . , . . .. . .... ... . . . . . . . ...... . ......... , , . , , . , , , . , , . , , , ... , ... , . . .. . . .... . . . . . . . .. ... .. . . . . . . . . . . .. . . .. . .. . . . . . . . . . .
' L = 1925
1934 'Feb
11. Another warning that the prior is over occurs thanks to the
harami pattern.
12. A dark-cloud cover. The letters X (from early February), Y (in
February), and Z (in late February) make up a three Buddha top.
13. A hammer.
14. Another harami. The prior short-term beginning at the
hammer (13) was short circuited with this harami.
15. Two windows which should act as resistance.
16. An inverted hammer. This also made a tweezers bottom.
17. The rally that started at the inverted hammer failed at the resistance
level made by an open window in (15).
18. A harami then hinted that the downtrend was over.
19. A dark-cloud cover.
9400
. 9012
9018 Jun Ju
.
THE RULE OF
MULTIPLE TECHNICAL
TECHNIQUES
1. A 50% retracement of the 1987 high (Area A) at $502 and the 1989
lows (1and 2) at $357 was $430.
2. The lows, marked 1 and 2 in 1989, were a double bottom. Based on
this double bottom, I derived a measured target of $425. (A double
bottom measured move is derived by taking the height of the move
between the two lows and adding this distance to the intervening
high).
3. The late 1988 high was $433.
4. My colleague, John Gambino, who follows Elliott Wave Theory,
said gold was in an Elliott fourth wave count. Based on this, rallies
in gold should not pierce the prior first wave's low in early 1988
at $425.
1. This was a multi-tested old support level from early March through
April. I believed that once this strong support broke, it should
become equally strong resistance.
2. The 65-day moving average (which I find useful for many markets)
The Rule of Technical Techniques 175
...............................................................................
. . . . . . . . . . . . . . . .
.
A
. . . . .
' L
Day ........................................
Moving
Average
1
. :. . . . : . . .
I I
. : : :
1378
1388
1378
1380 'Jan 'Feb 'Mar 'May
12012714 11211912612
Notes
CONFLUENCE
OF CANDLESTICKS
a hanging man;
a doji;
and a shooting star which was the coup de'grace.
Doji
'
Man
I
I
Jun
Source.
of a bearish shooting star. This line's small real body (being within the
previous day's real body) makes it a harami. Finally, the top of the upper
shadow (that is, the high of the day) on the shooting star day was also
a failure at the February 1600 highs.
Exhibit 10.4 shows that within a period of a few weeks, this market
formed a tweezers bottom, a bullish engulfing pattern, and a hammer.
Exhibit 10.5 shows that from mid to late July, a series of bearish candle-
stick indications occurred including a doji star followed by three
hanging-man lines (as shown by 1, 2 and 3). In between hanging-man 1
and 2, a shooting star formed.
Exhibit 10.6 is a bearish candlestick signal within a bearish candle-
stick signal. The peak of the rally in December was touched by a
hanging-man session. This hanging-man session was also the star por-
tion of an evening star formation. Exhibit 10.7 shows that May 9 through
11 delivered a series of top reversal candlestick signals at the $1.12 area.
The tall white candlestick on May 9 was followed by a small real body
candlestick. This second candlestick was a hanging man. It also, when
A Confluence of Candlesticks 179
Shooting Star
and
Pattern
Resistance
'3
11
4'9
I I
I
Tweezers Bottom
I I I I
Source: CompuTracTM
. . . . . . Hanging
. . . . . . .
. . . . . . . . . Evening Pattern
. .
. .
. .
. .
. . .
I . .
. . . .
November December
12000
.
. . . . .
.. ..
. . . ,
.
..
..
..
..
11000
, .
7i . . ..
11650
11885 , . . , .
. 11640
11820 . Jun
(2
a hanging man;
a harami;
a shooting star; and
tweezers top.
The market backed off from these highs. The $1.12 price became sig-
nificant resistance as evidenced by the bulls' failure to punch above it
during mid-June's rally. This $1.12 level was important for another rea-
son. Once broken on the upside on June 28, it converted to pivotal sup-
port. Observe the doji star that arose after the June 28 long white
candlestick. We know a doji after a long white candlestick is a top rever-
sal. This means the prior should end. For two days after the
doji, the market showed it was running out of breath since there were
two black candlesticks locked in a lateral band. The market had run out
of steam-or so it had appeared. Remember the May 9 through 11 resis-
tance area? The lows of the two black candlestick sessions of July 2 and
3 held that old resistance as support. The bears had tried to break the
A Confluence of Candlesticks 183
market but they could not. Until that support broke, the back of the
short-term bull market which commenced June 26 would not be broken.
In this scenario, the confluence of candlesticks which was so important
as a top on May 9 through 11 became influential again a few months
later as important support.
Exhibit 10.8 illustrates how, in mid-1987, a series of candlestick sig-
nals intimated a top. Specifically, within a month there was a hanging
man, a doji, and a dark-cloud cover. After the dark-cloud cover, the
market sold off and, in the process, opened a window. This window
became resistance on the brief rally just before the next leg lower. The
finally ended with the tweezers bottom and the bullish belt-hold
line (although the white candlestick had a lower shadow it was small
enough to view this line as a bullish belt hold).
CHAPTER
CANDLESTICKS
WITH TRENDLINES
. . . . . . . . . . . . . . . . . .
I I I I I
' 170
23 20
Line .
3/21/90
2016
,,,,,, .... , , , ... ........................ ,. , , , , ,. ... ..... ........... , .. , .. . . . ..... ........... ... . .... . . . ..... . . . . . . . . . . . . ......
, , ,
1900
2004
2013 'Jan 'Feb 'Mar
EXHIBIT 11.5. Crude Oil-June 1990, Daily (Support Line with Candlesticks)
ket pulled back. You should be confident when the trade is placed, but
always take into account doubt and uncertainty. One of the most impor-
tant concepts in futures, is risk control. The use of
stops is synonymous to risk control.
Exhibit 11.6 shows dark-cloud covers 1 and 2 produced a resistance
line. Dark-cloud cover 3 intersected at this resistance line and thus con-
firmed this line's importance as a supply area. Exhibit 11.7 shows that
there was a rally (not shown) that stopped at A. This area provided a
preliminary resistance area at A long-legged doji arose at B. The
fact that this doji also surfaced near the resistance level set by A was a
reason to be cautious. Points A and B gave the first two points of a resis-
tance line. Traders who use hourly charts would thus for failed ral-
lies near this line to take appropriate action-especially if they got a
confirmatory bearish candlestick indicator. At C, there was a long-legged
doji (like the one at B) near the resistance line. The market then backed
off. At D, the white candlestick with a long upper shadow was a shoot-
ing star. It failed at the resistance line. This white candlestick was imme-
diately followed by a black candlestick that engulfed it. These two
candles constituted a bearish engulfing pattern.
Exhibit 11.8 shows two engulfing patterns where pattern 1 was a
Candlesticks with Trendlines 189
................................................................................
THE IMPORTANCE OF PROTECTIVE STOPS
Technicals should be used to set up parameters. As such they
will provide the analyst with a mechanism for a risk and money manage-
ment approach to trading. Defining risk means using protective stops to
help protect against unanticipated adverse price movements. If stops are
not used, the analyst is not taking advantage of one of the most powerful
aspects of technical analysis.
A stop should be placed at the time of the original trade; this is when
one is most objective. Stay in the position only if the market performs
according to expectations. If subsequent price action either contradicts or
fails to confirm these expectations, it is time to exit. If the market moves
opposite to the chosen position you may think, "why bother with a
stop-it is just a short-term move against me." Thus you stubbornly stay
with the position in the hope the market will turn in your direction.
Remember two facts:
The market does not care whether you own it or not. The one thing
worse than being wrong is staying wrong. Lose your opinion, not your
money. Be proud of the ability to catch mistakes early. Getting stopped
out concedes a mistake. People hate to admit mistakes since pride and
prestige get involved. Good traders will not hold views too firmly. It has
been said that famous private investor Warren Buffet has two rules:
Stops are synonymous with rule 1. You have limited resources. These
resources should be maximized, or at a minimum, preserved. If you are in
a market that has moved against your position, it is time to exit and find
a better opportunity. Think of a stop as a cost of doing business.
Since so much of the Japanese candlestick terminology is grounded on
military terminology, we will look at stops in this context as well. Each
trade you make is a battle. And you will have to do what even the great-
est generals have to do-make temporary, tactical retreats. A general's
goal is to preserve troops and munitions. Yours is to save capital and
equanimity. Sometimes you must lose a few battles to win the war. The
Japanese have a saying, "a hook's well lost to catch a salmon." If you are
stopped out, think of it as you would a lost hook. Maybe with the next
hook you will catch your prize.
190 The Rule of Multiple Technical Techniques
8000
.
,A= +32 ,
. . . . . . . .
................
. .
. .
. .
. . . . .
. . .
. . . .
.
. .
. . . .
. . .
. . .
. . . . .
. . . .
. .
. .
. . . . . . .
. . . . . .
I A/ 3 I 0 I
EXHIBIT 11.7. Japanese Yen-June, 1990 Intra-day (Resistance Line with Candlesticks)
Candlesticks with 191
warning to longs. A few weeks later, the second bearish engulfing pat-
tern emerged. The highs on engulfing pattern 2 also were a failure at a
resistance line. Exhibit 11.9 shows an upward sloping resistance line. It
is a trendline that connects a series of higher highs. While not as popu-
lar as the downward sloping resistance line in Exhibit 11.1, it can be a
useful device for longs. When the market approaches this kind of line,
longs should take defensive measures in anticipation of a pullback.
These protective measures could include taking some profits on long
positions, moving up a protective stop, or selling calls. Although pull-
backs should be temporary (since the major trend is up), the failure from
this line could be an early and very tentative indication of the beginning
of a new downtrend.
Exhibit 11.10 is a downward sloping support line. This is another
type of line not used very often, but can occasionally be valuable for
those who are short. the downward sloping support line is
192 The Rule of Multiple Technical Techniques
EXHIBIT 11.11. Cotton-May, 1990, Daily (Upward Sloping Resistance Line and Downward Sloping
Support Line with Candlesticks)
Price
Unsustained Objective
Unsustained
Price
New Lows
Objective
(Spring)
Upthrust Spring
or bulls may assault a prior high or low level. Trading opportunities can
arise on these occasions. Specifically, if there is an unsustained breakout
from either a support or resistance level, it can present an attractive
trading opportunity. In such a scenario there is a strong probability there
will be a return to the opposite side of the congestion band.
There is an unsustained penetration of resistance in Exhibit 11.12.
Prices then return back under the old highs which had been "pene-
trated." In such a scenario, one could short and place a stop above the
new high. The price target would be a retest of the lower end of the con-
gestion band. This type of false upside breakout is called an upthrust. If
an upthrust coincides with a bearish candlestick indicator it is an appeal-
ing opportunity to short.
The opposite of an upthrust is the spring. The spring develops when
prices pierce a prior low. Then prices spring back above the broken sup-
port area (see Exhibit 11.13). In other words, new lows could not hold.
Buy if prices push back above the old lows. The objective would be for
a retest of the congestion zone's upper band. The stop would be under
the lows made on the day of the spring. Trading springs and upthrusts
is so effective because they provide a clear target (the opposite end of the
trading range) and protective stop (the new high or low made with the
"false breakout").
Exhibit 11.14 is a good example of upthrusts with candlesticks. Day
A marked the high for the move and a resistance level (notice how the
hanging-man line the prior day gave warning of the end of the uptrend).
The dual lows at and defined the lower end of the trading band.
There was an upthrust on day B. That is, the prior highs at A were
breached, but the new highs did not hold. The failure of the bulls to
maintain the new highs at B was a bearish signal. Another negative sign
was that day B was also a shooting star. Shooting stars are sometimes
part of an upthrust. At such times, it is a powerful incentive to sell. As
if a bearish upthrust and a shooting star were not enough to send chills
down a back, the day after B a hanging man appeared! With the
Candlesticks with Trendlines 195
. . .
.......................................
26000
. . .
'Feb
120 119
. . .
. . .
.. . ,
.. . . .
..
.
,
'Nay 'Jun
(9
Mar
Exhibit 11.18 indicates that the early January lows were perforated in
late February. The failure to hold the lows meant that this was a bullish
spring. The day of the spring was also a hammer. This union of bullish
signals gave plenty of warning to the technician to look for a return
move to the upper end of the band near $78. Interest-
ingly, the rally stopped in mid-March near $78 at an evening doji star
formation.
Exhibit 11.19 shows that after a harami, the market slid. It stabilized
at hammer 1. This hammer was also a successful test of the prior sup-
port near Another slight pullback occurred on hammer 2. With this
bullish hammer, the market nudged marginally under the summer lows
(by 25 ticks) but the bears could not maintain these new lows. Thus a
spring, complimented by a hammer and a tweezers bottom created note-
worthy bullish evidence. Exhibit 11.20 reveals that during the week of
March 12, soybeans touched a low of $5.96 formed a bullish engulfing
198 The Rule of Multiple Technical Techniques
. .....................................................................
and
Hammer
Source: L.P.
A= ......................................................................................................................................
Hararni
7/23/90
Hammer 1 2
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
., .
.
Pattern
Engulfing..
.
3/12 3/19 3/26 7 9 4/16 4/72
pattern and rallied. On April 3, prices broke this level and made new
lows. These new lows failed to hold and created a spring. Furthermore,
the lows on that session constructed a bullish engulfing pattern.
Why do springs and upthrusts work so well? To answer this, refer to
Napoleon's response when asked which troops he considered best. His
terse response was, "those which are View the market as a
battlefield between two sets of troops-the bulls and the bears. The ter-
ritory they each claim is especially evident when there is a lateral trad-
ing range. The horizontal resistance line is the bears' terrain to defend.
The horizontal support line is the bulls domain to defend.
1
3522
3476
3482 'May 'Jun
1
The Japanese have a saying that, "a red lacquer dish needs no decora-
tion." This concept of simple beauty is the essence of a technical princi-
ple I frequently use with candlestick charting. It is as simple as it is
powerful-old support becomes new resistance; old resistance becomes
support. This is what I call the "change of polarity" principle. Exhibit
11.22 shows support converting to resistance. Exhibit 11.23 illustrates
prior resistance becoming new support. The potency of this change of
polarity is proportional to:
Converted to Old
Resistance
Converted t o
New Support
Resistance
4/16/90
, 6280
6420
6254
6365 t 'Jan Ju 'Jan Ju 'Jan
11988
. . . . 7100
. .
. . . . .
.
. ..
. 7/30 . . . . . . .
7325 . . . . . . .
7350 . . . . . .
7324 7000
. 7348 . . .
7/16
a dotted line near $1.85. When this level was breached in June 1990,
what was next support? The price action from $1.25 to $2.05 was essen-
tially straight up so there was no support evident based on the late 1989
to early 1990 rally. Yet, when $1.85 broke, support was expected near
$1.65. Where did I get that figure? Two reasons. The first was that a 50%
correction of the prior rally was near $1.65. The second, and more
important reason, was the prior resistance at area A was also near $1.65.
That should mean it will now be support. A series of limit-down days
comprised the June This stopped at $1.66.
Pick up just about any chart, be it intra-day, daily, weekly, or longer
and the chances are high that you will see examples of this change of
polarity in action. Why is something so simple so good? The reason has
to do with the raison of technical analysis; to measure the emo-
tions and actions of the trading and investing community. Thus, the bet-
ter a technical tool measures behavior, the better that tool should work.
And the change of polarity principle is so successful because it is based
on sound trading psychological principles. What are these principles? It
has to do with how people react when the market goes against their
position or when they believe they may miss a market move.
Ask yourself what is the most important price on any chart? Is the
Candlesticks with Trendlines 205
highs made for move? The lows? Yesterday's close? No. The most impor-
tant price on any chart is the price at which you entered the market. People
become strongly, keenly and emotionally attached to the price at which they
bought or sold.
Consequently, the more trading that transpires at a certain price area
the more people are emotionally committed to that level. What does this
have to do with the fact that old resistance becomes support and old
support becomes resistance? Let us look at the Exhibit 11.29 to answer
this. In late December, a steep culminated at $5.33 (at A). On
another test of this level, there are at least three groups who would con-
sider buying.
Group 1 would be those who were waiting for the market to stabilize
after the prior and who now have a point at which the market
found support-$5.33 (the December 28 lows at area A). A few days
later, a successful test (at B) of this support probably pulled in new
longs.
Group 2 would be those who were previously long but were stopped
out during the late December On rally B to in mid-January,
some of these old longs who were stopped out would say to themselves
that they were right about silver being in a bull market. They just timed
there original purchase incorrectly. Now is the time to buy. They want
to be vindicated in their original view. They wait for a pullback to sup-
port at C to go long again.
Group 3 would be those who bought at points A and B. They also see
the B to rally and may want to add to their position if they get a "good
price." At area C, they have their good price since the market is at sup-
port. Thus more buyers come in at C. Then for good measure another
pullback to D draws in more longs.
Then the problems start for the longs. In late February, prices punc-
ture support areas A, B, C, and D. Anyone who bought at this old sup-
port area is now in a losing trade. They will want to get out of their trade
with the least damage. Rallies to where the longs bought (around
will be gratefully used by them to exit their longs. Thus, the original
buyers at areas A, B, C, and D may now become sellers. This is the main
reason why old support becomes new resistance.
Those who decided not to liquidate their losing long positions on the
minor rallies in early March then had to go through the pain of watch-
ing the market fall to $5. They used the next rally, in early April (Area
E), to exit. Exhibit 11.29 illustrates how support can become resistance.
The same rationale, but in reverse, is the reason why resistance often
becomes support. Do not let the simplicity of the rule fool you. It
works-especially when melded with candlestick indicators. For exam-
ple look at area E. Notice how the doji after a tall white real body meant
206 The Rule of Multiple Technical Techniques
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
t . : . .:.
. . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . .
.
.: 5500
. . Resistance
. . . .
. . . . .
.
: 5000
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . : . . . . . . .:
. . .
4815
trouble. This candlestick signal coincided with the resistance line. The
same scenario unfolds at F.
In Exhibit 11.30, the highs at A and B then became support in late
1986 and then in 1989. Note how the strength of this support was con-
firmed twice in 1989 by two consecutive hammer lines.
In September and early October, at areas and in Exhibit 11.31,
the market maintained a support level near $1,230. Once the bears
pulled the market under that level on October 9, this $1,230 then con-
verted to a band of resistance. After the first failure at this new resis-
tance, at C, prices descended until the bullish engulfing pattern. A
minor rally then followed. This rally stalled, once again, at the $1,230
level. In addition, there was a dark-cloud cover.
Candlesticks with Trendlines 207
Dark
Cloud
Cover
Bullish
Engulfing
............................................................................................................
1165
'H= 1165
'L= 1150
1150 'Oc t
Notes
CANDLESTICKS WITH
RETRACEMENT LEVELS
Markets usually do not trend straight up, nor do they fall vertically
downward. They usually retrace some of the advance, or decline, before
resuming the prior trend. Some of the more popular retracement levels
are the 50% level and the Fibonacci figures of 38% and 62% (see Exhib-
its 12.1 and 12.2). Fibonacci was a 13th century mathematician who
derived a special sequence of numbers. Without getting into too much
detail, by comparing these numbers to one another one could derive
what is called-not surprisingly-the Fibonacci ratios. These ratios include
61.8% (or its inverse of 1.618) and 38.2% (or its inverse of 2.618). This is
why the 62% (61.8% rounded off) and the 38% (38.2% rounded off) cor-
rections are so popular. The popular 50% correction is also a Fibonacci
ratio. The 50% retracement is probably the most widely monitored level.
This is because the 50% retracement is used by those who use Gann,
Elliott Wave, or Dow Theory.
Exhibit 12.3 illustrates how well retracements can help predict resis-
tance areas in a bear market. The 50% retracements in gold over the past
few years have become significant resistance levels. Let us look at three
instances on this chart where 50% retracements melded with candlestick
techniques to provide important top reversal signals.
1-The highs at A in late 1987 ($502) were made by a
bearish engulfing pattern. The which began in late 1987 ended
with a piercing pattern at B at $425. Based on a 50% retracement of this
from A to B, resistance should occur at $464 (this is figured by
taking half the difference between the high at A and the low at B and
The Rule of Multiple Technical Techniques
B
- 3 8 % Retracernent of A- B
- 5 0 % Retracernent of A-B
- 6 2 % Retracement of A-B
- 3 8 % Retracement of A-B
- 5 0 % Retracernent of A- B
- 6 2 % Retracernent of A-B
\
B
EXHIBIT 12.2. Popular Retracement Levels
in a Downtrend
adding this onto the low at B). Thus, at $464 you look for resistance and
confirmation of resistance with a bearish candlestick indicator. A bearish
engulfing pattern formed at C. At C, the high was $469 or within $5 of
the 50% correction. The market began its next leg lower.
Retracement 2-The which began at C ended at the morning
star pattern at D. Taking a 50% correction from high at $469 to
low at $392 gives a resistance area of $430. Thus, at that level, bearish
candlestick confirmation should appear. Gold reached $433 at area E.
During this time (the weeks of November 28 and December 5 (at E)) gold
came within of making a bearish engulfing pattern. Another decline
started from E.
Retracement 3-From the high at E to the low at F in 1989 (at
prices fell $76. (Interestingly, all three selloffs, A to B, C to D, and E to
F fell about $77.) There were no candlestick indicators that called the
lows on June 5. The second test of these lows in September came via a
hammerlike line.
The next resistance level, a 50% retracement of the decline from E to
F, is $395. Not too surprisingly gold surpassed this level. Why wasn't
this a surprise? Because, in late 1989, gold pierced a two-year resistance
Candlesticks with Retracement Levels
line. In addition, gold built a solid base in 1989 by forming a double bot-
tom at the $357 level. Thus, we have to look out farther to a 50%
ment of the larger move. This means a 50% retracement of the entire
decline from the 1987 high (area A) to the 1989 low (area F). This fur-
nishes a resistance level of $430. Near this $430 level, at $425 on the
week of November 20 at area G, the market gave two signs that the
was in trouble. Those signs were a harami pattern and, as part
of this pattern, a hanging man. A few weeks later, on the week of Janu-
ary 22, the highs for this move were touched at $425. The following
week's price action created another hanging man. Gold declined from
there.
Look at Exhibit 12.4. The combination tweezers and harami bottom at
$18.58 (A) was the start of a $3.50 rally. This rally terminated at $22.15
(B) with a bearish engulfing pattern. A 50% correction of the A-B thrust
would mean support near $20.36. At area C, a bullish piercing pattern
formed at $20.15. The market then had a minor rally from C. This rally
ran into problems because of the dark-cloud cover at D. Interestingly,
high at $21.25 was within 10 ticks of a 50% bounce from the prior
12 44 CLKO D A I L Y BAR
. 1836
1845 (Bearish . . . . . .
. 1752
Engulfing
D
,
Pattern)
. . .
. . . . . . . . . . . . . . . . . . . . . .
and Hararni)
1752
1795 'Nov 'Jan 'Feb 'Mar
(30
. . . . . .
. . . . . .
. . . . . .
. . . . . . ..
Hararni . ..
. . ..
2/90
, 6430 .
6530
6516 Mar ,
19 26 2
-39 ,
. . . : . . . : :
. . . . . . . . . . . . . . . . Hammer .
3000
EXHIBIT 12.6. Crude Oil-(a) December 1990 and (b) December 1990 Intra-day
214 The Rule of Multiple Technical Techniques
$5.97. This also coincides closely with the old resistance level from late
January and February at $5.95. That old resistance converted to support.
On the pullback to this level, on April 2 and 3, this $5.97 held as sup-
port. These sessions formed a harami pattern that signaled an end to the
prior minor downmove. Then, just for good measure, there was an addi-
tional test of this support in mid-April, and away went the beans!
Exhibit shows from crude oil's July low at L, to its October high
at H, there was a $21.70 rally. A 50% retracement of this rally would be
at $29.05. Thus, based on the theory that a 50% retracement level from
a rally should be support, we should look for a bullish candlestick indi-
cator near that $29.05 area on the brisk from October's high. This
is what unfolded. On October 23, after prices touched a low of $28.30, a
hammer developed on the daily chart. The market rallied over $5 from
that hammer. On the intra-day chart of the price action on October 23
(see Exhibit we see the first hour's action also formed a hammer.
Thus, the daily candlestick chart on October 23 and the first hour on the
intra-day candlestick chart on October 23 both had hammers. This is a
rare, and as we see, significant occurrence. Note how, on the intra-day
chart, the brisk rally that began with the hammer ran out of force with
the emergence of the hanging man on October 26.
CANDLESTICKS WITH
MOVING AVERAGES
T h e moving average is one of the oldest and most popular tools used by
technicians. Its strength is as a trend-following device which offers the
technician the ability to catch major moves. Thus, it is utilized most
effectively in trending markets. However, since moving averages are
lagging indicators they can catch a trend only after it has turned.
The examples that follow use various moving averages. They are not
based on optimum moving averages. An optimum moving average
today might not be the optimum one tomorrow. The moving averages
used in this text are widely monitored along with some which are not as
widely used but which are based on such tools as Fibonacci numbers.
The moving averages used here are not the important point. What is
meaningful is how moving averages can be melded with candlesticks.
I like using the 65-day moving average as a broad spectrum moving
average. From my experience, it seems to work well in of the
futures markets. Exhibit 13.1 illustrates a 65-day moving average that
offered support to the market at areas 1, 2, and 3. Beside the moving
average shoring up the market at these points, we see a bullish engulf-
ing pattern at area 1, a hammer and harami at 2, and another
like line at 3.
Exhibit 13.2 reveals that a confluence of technical factors joined on
April 2 and 3 to warn alert eyes of trouble ahead. Let us take a look at
the specifics:
1. In early March, prices broke under the 65-day moving average. From
that point, the moving average became resistance.
. . . . . .
. . . . . .
Average
.
6510 .
6520 6055.91...................................................................................................................
6452
6464 Mar
EXHIBIT 13.1. Soybeans-July 1990, Daily (Simple Moving Average with Candlesticks)
Candlesticks with Moving Averages 219
..........................
EXHIBIT 13.2. Crude Oil-June 1990, Daily (Simple Moving Average with Candlesticks)
Exhibit 13.3 shows that late February's test of the 65-day moving
average support line was confirmed with a hammer. The market retested
these lows a few days later and, in the process, formed a tweezers bot-
tom.
The Rule of Multiple Technical Techniques
65 SIM
............ ....................................... ....
...
. .
Hammer
. . Moving . . . .
Average
EXHIBIT 13.3. Sugar-May 1990, Daily (Simple Moving Average with Candlesticks)
...................... ......
. . . . . .
....
. . .
'I
I. ....................................... I I. . I.
:
. . .
.
EXHIBIT 13.4. Copper-September 1990, Daily (Dual Moving Averages with Candlesticks)
The Rule of Multiple Technical Techniques
.:........................................
.:.
. . .
EXHIBIT 13.5. Deutschemark-September 1990, Daily (Dual Moving Averages with Candlesticks)
Dark-cloud
.....
. . .
. . .
1890
1909 . . . . . .
1830
1831
.......................................................
. . .
......
.. .. ..
1= . . . . . .
. 126
EXHIBIT 13.6. Crude Oil-June 1990, Daily (Dual Moving Averages with Candlesticks)
..
9000
17/90
8917 . . . . .
8924
8902
'C= 8914
OSCILLATOR BAR .
0.150 EXP - 0.075 EX P
.. . .
:
The MACD has two lines. They are shown on the lower chart in
Exhibit 13.7. The more volatile solid line is the signal line. A sell signal
occurs when this signal line crosses below the dashed, less volatile line.
In this example, the bearish implications of the two bearish engulfing
patterns were corroborated by the bearish crossovers of the MACD indi-
cators (see arrows).
In Exhibit 13.8, the signal line of the MACD pushed above the slower
moving line in early July (see arrow). This was a notable clue that the
market might be bottoming. Shifting to the candlesticks shows that the
first morning star's bullish implications were voided by the dark-cloud
cover. The price decline from this dark-cloud cover ended with another
morning star. After a temporary set back with the hanging man, the
market's upward force gained steam.
CANDLESTICKS
WITH OSCILLATORS
if
"Let every bird sing its own note"
OSCILLATORS
Oscillators include such technical tools as the relative strength index,
stochastics, and momentum.
As discussed in greater depth later in this chapter, oscillators can
serve traders in at least three ways:
RSI = 100 -
Thus, computing a 14-day RSI entails adding the total gains made on
the up days over the last 14 days (on a close-to-close basis) and dividing
by 14. The same would be done for the down days. These figures pro-
vide the relative strength, (RS). This RS is then put into the RSI formula.
This RSI formula converts the RS data so that it becomes an index with
a range between and 100.
. , .
I
.
. , . .
Bullish
., Engulfing
.. .. HSI CPKO DAILY : . . .
. . ., . . . . .
I
.
. . . .. .. .. :
.. .. .. . .
..
..
.. . . . .. .. .
I. I. I.
. 30001
., Hararni
Shooting
27001
.. . . .
STOCHASTICS
-
(High of - (Low of
27000
..
.. .. . .
. .
EXHIBIT 14.4. Dow Jones Industrial Average, 1989-1990 (Stochastics with Candlesticks)
There was also a positive crossover as the more volatile, solid line
crossed above the dotted line (see arrow). This crossover is consid-
ered more significant if it is from oversold readings (that is, under 25%).
That is what occurred here. So, in early 1990, there were a series of
hammerlike lines and a positive divergence with a positive crossover
during an oversold market. A confluence of technical indicators that
were strong clues that the prior downtrend had ended.
As illustrated by Exhibit 14.6, April 12 and 16 formed a dark-cloud
cover. The black candlestick session on April 16 pushed prices above the
former highs in March. Thus prices were at a new high, but stochastics
were not. The dark-cloud cover and the negative divergence were two
signs to be circumspect about further rallies. The next was cor-
roborated by a negative crossover when the faster line crossed under
the slower line (as shown by the arrow).
I do not often use candlesticks with British Pound futures because, as
can be seen in Exhibit 14.7, many sessions are small real bodies or doji.
This is in addition to the frequent gaps induced by overnight trading
(this is also true of other currency futures). Nonetheless, at times, there
are candlestick signals that bear watching especially when confirmed
with other
.
Dark-cloud
A= +35
............
. . .
.
. . .
32925
. 32975
32920
SLOU STOCHASTIC SPMO DAILY BAR
.. .. .. ................. 75
. . .
50
. .
.:. .:. ........
.. .. .
. . .
. . .
. . .
. . . . . .
. 3/28/70. 15500
.. . . . Star .
.. . . . . . .
,
. .
.
.. .. .. .. . . .
.. . . . ...
.. .. . . 25
63.41
-D= 41.80
112
MOMENTUM
Notes
'This RSI is different than the relative strength used by equity technicians. The relative strength
used by equity technicians compares the relative strength performance of a stock, or a small
group of stocks, to the performance of broader market index such as the Dow Jones or the
500.
help follow the 24-hour foreign exchange markets some Japanese candlestick users will draw
a candlestick based on the Tokyo trading session and draw another candlestick line on the U.S.
trading session. Thus, for each 24-hour period, there will be two candlesticks. For those who fol-
low the currency futures, the weekly candlestick charts may decrease some of the problems
caused by overnight trading.
CANDLESTICKS WITH
VOLUME AND OPEN
INTEREST
T h e theory behind volume states that the greater the volume, the greater
the force behind the move. As long as volume increases, the current
price trend should continue. If, however, volume declines as a price
trend progresses, there is less reason to believe that the trend will con-
tinue. Volume can also be useful for confirming tops and bottoms. A
light volume test of a support level suggests a diminution of selling force
and is, consequently, bullish. Conversely, a light volume test of a previ-
ous high is bearish since it demonstrates a draining of buying power.
Although volume can be a useful auxiliary medium to measure the
intensity of a price move, there are some factors with volume, especially
as they pertain to futures, that somewhat limit their usefulness. Volume
is reported a day late. Spread trading may cause aberrations in volume
figures-especially on individual contract months. With the increasing
dominance of options in many futures markets, volume figures could be
skewed because of option arbitrage strategies. Nonetheless, volume
analysis can be a useful tool. This chapter examines some ways volume
and candlestick charting techniques can be merged.
242 The Rule of Multiple Technical Techniques
Exhibit 15.1 shows how volume and candlestick techniques can help
confirm double tops or bottoms. On March 22 (line the market
pushed up to the late February highs near 94. Volume at line 1 was
504,000 contracts (all volume figures are total volume for all contract
months). For the next few sessions, prices tried to push above this 94
level. The small white real bodies on these days reflected the bulls' lack
of fervor. The low volume figures on these small candlestick sessions
echoed this. The bulls finally surrendered after a week. In late March,
within two days, the market fell two full points.
Next we turn our attention to the tall white candlestick of April 4 (line
2). Could this strong session presage gathering strength by the bulls?
The answer is probably not. First, we note the volume on this rally ses-
sion was a relatively light 300,000 contracts. The long black candlesticks
a few sessions before (March 29 and 30) had larger volume. Another sign
of trouble appeared with the action following line 2. The next day (line
3) a small real body appeared. Lines 2 and 3 constituted a harami pat-
tern. The implications were that the prior was over. Note also
that this small real body day was also a variation on a bearish hanging
ED DAILY , .. . .
..
..
.:. ,:. ..
. .
ON BALANCE VOLUME
On balance volume (OBV)is a net volume figure. When the market closes
higher compared to the prior close the volume figure for that day is
added to the cumulative on balance volume figure. When the close is
lower, the volume for that day is subtracted from the cumulative on bal-
ance volume figure.
OBV can be used in a few ways. One way is to confirm a trend. OBV
should be moving in the direction of the prevailing price trend. If prices
are ascending along with OBV, increased volume is reflected by the buy-
ers, even at higher price levels. This would be bullish. If, conversely,
price and OBV are declining, it reflects stronger volume from the sellers
and lower prices should continue.
OBV is also used in lateral price ranges. If OBV escalates and prices
are stable (preferably at a low price area) it would exhibit a period of
accumulation. This would bode well for advancing prices. If prices are
moving sideways and OBV is declining it reflects distribution. This
would have bearish implications, especially at high price levels.
Candlesticks with Volume and Open Interest 245
TICK VOLUME
The hourly intra-day chart in Exhibit 15.4 shows the usefulness of tick
volumeTM. After a bullish hammer late in the session on May 4, prices
moved higher. However, these advancing prices were made on declin-
ing tick volumeTM. This was one sign of lack of conviction by buyers. The
other was the short white real bodies. In the first three hours of May 8,
there was a sharp price break. These made new lows for the move. The
intra-day action late on May 8 provided clues this early morning
was to be short lived. After the third hour's long black line, a doji mate-
rialized. These two lines formed a harami cross. Then a white body
appeared a few hours later which engulfed the prior two black bodies.
This was a bullish engulfing pattern that had extra significance since it
engulfed two black bodies. The lows made by the white engulfing line
also formed a tweezers bottom.
Just in case, another hint of a bottom was needed, tick volume " sub- T
stantiated that the buyers were taking control. Prices rose after the afore-
mentioned bullish engulfing pattern. During this rally, volume
expanded as did the height of the real bodies. A shooting star and resis-
tance near $1,340 from the prior week, temporarily put a damper on the
price ascent. Once the market pushed above the $1,340 resistance level
via a window, there was no doubt the bulls were in control.
In Exhibit 15.5, the hammer hour on June 19 furnished a sign that the
market may be searching for a bottom. The first hour of June 20 made a
new low at $16.62 (line 1). Tick volumeTM on this hour was a brisk 324
trades. Another move down to that level, via a long black candlestick
(line 2), was made later that session. This time tick volumeT" was only
262 trades. The next session, June 21, is the one of the most interest. On
the third hour of trading, prices made a new low for the move at $16.57.
This new low was made on lighter tick volumeT" (249 trades) then the
prior two tests (lines 1 and 2). This meant selling pressure was easing.
Prices then sprung back and made an hourly hammer line. (From the
EXHIBIT 15.5. Crude Oil-August 1990, Intra-day (Tick VolumeTM with Candlesticks)
248 The Rule of Multiple Technical Techniques
OPEN INTEREST
In the futures markets, a new contract is created when a new buyer and
a new short seller agree to a trade. Because of this, the number of con-
tracts traded in the futures market can be greater than the supply of the
commodity which underlies that futures contract. Open interest is the
total number of long or short contracts, but not a total of both, which
remain outstanding.
Open interest assists in gauging, as does volume, the pressure
behind a price move. It does this by measuring if money is entering or
exiting the market. Whether open interest rises or falls is contingent on
the amount of new buyers or sellers entering the market as compared to
old traders departing.
In this section, our focus will be on the importance of price trends
accompanied by rising open interest. The major principle to keep in
mind is that open interest helps confirm the current trend if open inter-
est increases. For example, if the market is trending higher and open
interest is rising, new longs are more aggressive than the new shorts.
Rising open interest indicates that both new longs and new shorts are
entering the market, but the new longs are the more aggressive. This is
because the new longs are continuing to buy in spite of rising prices.
A scenario such as building open interest and falling prices reflects
the determination of the bears. This is because rising open interest
means new longs and shorts entering the market, but the new short sell-
ers are willing to sell at increasingly lower price levels. Thus, when open
interest rises in an the bulls are generally in charge and the
rally should continue. When open interest increases in a bear trend, the
bears are in control and the selling pressure should continue.
On the opposite side, if open interest declines during a trending
market it sends a signal the trend may not continue. Why? Because for
open interest to decline traders with existing positions must be abandon-
ing the market. In theory, once these old positions are exited, the driv-
ing force behind the move will evaporate. In this regard, if the market
rallies while open interest declines, the rally is due to short covering
(and old bulls liquidating). Once the old shorts have fled the market, the
force behind the buying (that is, short covering) should mean the mar-
ket is vulnerable to further weakness.
As an analogy, let's say that there is a hose attached to a main water
Candlesticks with Volume and Open Interest 249
line. The water line to the hose can be shut off by a spigot. Rising open
interest is like fresh water pumped from the main water line into the
hose. This water will continue to stream out of the hose while the spigot
remains open (comparable to rising open interest pushing prices higher
or lower). Declining open interest is like closing the spigot. Water will
continue to flow out of the hose (because there is still some water in it),
but once that water trickles out, there is no new source to maintain the
flow. The flow of water (that is, the trend) should dry up.
There are other factors to bear in mind (such as seasonality), which
we have not touched upon in this brief review of open interest.
7/12/90 . .. ..
. .. ..
. 7355
7207 . .. ..
7349
.. CT DAILY .
a
"Like both wheels of a cart"
ratios go hand in hand since these ratios can be used to project price tar-
gets for the next wave. Thus, for example, wave 3 could be projected to
move 1.618 times the height of wave 1; a wave 4 could correct 38.2% or
50% of the wave 3 move; and so on.
. . . . ..
Dark-cloud
Cover
Cover , .. ,
Harami
and
Shooting Tweezers
.. Top
Harami
. . .
: :
Piercing Morning
. Star
. . . .
5/23/90
.
.
EXHIBIT 16.2. Crude Oil-June 1990: (a) Daily and (b) Intra-day (Elliott Wave with Candlesticks)
256 The Rule of Multiple Technical Techniques
series of strong long white bodies. On this hourly chart the bullish ham-
mer and the ensuing white lines confirmed a bottom for the wave 3
count. The daily chart also provided a bottom reversal signal with a
piercing pattern based on the price action of April 10 and 11. We now
look for a wave 4 rally.
Corrective wave 4's top should rally but should not move above the
bottom of the prior wave 1 according to Elliott Wave theory. In this
example, that would be the $19.95. Approaching that level, look for a
bearish candlestick cue. That is, indeed, what unfolded. On May 14 and
15, the market stalled under $20 via the harami cross and tweezers top.
The wave 4 top was hit.
Interestingly, by breaking wave 4 into its (a), (b), and (c) subcompo-
n e n t ~ ,corroboration by candlestick indicators is apparent. At (a), a
shooting star whose highs stopped at a window from early April
appears. A morning star at (b) called the bottom. Subcomponent (c) was
also the top of wave 4 with its attendant bearish harami cross and twee-
zers top.
Exhibit 16.3 illustrates that the five impulse waves down began in
late December 1989 from 100.16. The wave counts are shown. The bot-
tom of wave 1 formed a harami pattern. The top of wave 2 formed a
, ,
9406 . Doji and Tweezers
9414 Jan Feb Mar Jun
122
Selling Extreme
Initial
Balance
Selling . Area
Range .
Extension
Buying Extreme
A price probe is the market's search for the boundaries of value. How
the trading and investing community acts on such price probes can send
out important information about the market to Market users.
One of two actions occur after a price probe. Prices can backtrack to the
value area or value can relocate to the new price. Acceptance of a new
price as value would be confirmed by increased volume and the time
spent at that level.
If prices backtrack to value, the market shows a rejection of those
prices that are considered unfairly high or low. Quick rejection of a price
can result in what is called an extreme. An extreme is defined as two or
more single at the top or bottom of the profile (except for the last
half hour). Normally, an extreme at the top of the profile is caused by
competition among sellers who were attracted by higher prices and a
lack of buyers. A bottom extreme is caused by an influx of buyers
attracted by lower prices and a dearth of sellers. Buying and selling
extremes are noted in Exhibit 17.1.
How the market trades compare to the prior value area also discloses
Q
Exhibit 17.2 reveals how a doji on July 2 and a hanging man during the
next session were candlestick alerts of a top. What did the Market Pro-
262 The Rule of Multiple Technical Techniques
-24 I
file@say during this time period? July 2 had a relatively small value area
as compared to the prior session. It was also a light volume session
(132,000 contracts as compared to 303,000 the prior session). This hinted
that price was having trouble being accepted at these higher rates. In
other words, there was lack of trade facilitation. It is also a profile with
range extension to both sides. This indicates a tug of war between the
bulls and bears.
July 3 was another light-volume session (109,000 contracts) which
significantly discounts the bullish developments-an upside range
extension and a close at the highs of the day. The next day, July 5, is
where the weakness of the market materializes. During the early part of
the session, new highs were made for the move. In the process, upside
range extensions were also formed. With these range extensions, the
market was advertising for sellers. They got them. The market sold off
toward the latter part of the session (the J, K, and L periods) to close
near the low of the day. The open of July 6 saw initiating selling on the
opening since the market opened under the prior day's value area. This
showed immediate selling activity. July 6 also displayed increased vol-
ume and an initiating selling extreme (that is, single prints at the top of
the profile) during the "y" period. This confirmed the market was in
trouble.
In this example, we see an important aspect about the hanging man
previously addressed; it is only what happens after that line that makes
it a bearish indicator. July 3 was the hanging-man day. Here we see the
Market Profile@picture of the hanging-man line was giving some posi-
tive indications about the market. It was only on the following session,
July 5 and especially the morning of July 6, that a top was also verified
via Market ProfileB.
Exhibit 17.3 shows that July 5 was a very evident shooting star. After
this line appeared, cotton plunged for three sessions. Was there any-
thing prior to the shooting star a Market Profile@which gave signs of
trouble? Yes, there was. From June 29 to July 3 prices advanced but by
way of a shrinking value area. This meant less trade facilitation at higher
prices. The market was having trouble accepting these new highs as
value. In addition, volume, as gauged by the total number of those ses-
sion's TPO counts, was decreasing (actual volume was light on these
sessions. However, since volume is not released until the day,
the TPO count can be used as a gauge of volume). Notice of a top was
provided by the shooting star of July 5. The Market Profile@on that ses-
sion showed a range toward the upside failed to get buying
through. Additionally, the entry of sellers attracted by these higher
prices drove the market down as indicated by the range extension down
and the weak close. These were bearish signs.
264 The of Multiple Technical Techniques
.. ..
..
..
I... ..
................... :
. .
..
. ..
: ..
t
CTZO 1990
7700 . 7588
7686 . FG 7588
7672 . EFGJKL 7509
7658 . DEFGHIJKL .
7644 . DEGHIJL
7630 . . .
7616 . DGL
7602 . DL
7588 . DL
7574 .
7560 . KL
7546 . DEFGHIJK .
7532 . DEFGHIJK .
7518 .
7504 .
7490 .
7476 .
Shooting
Star . .
Session . DEFGH . E
DEFGHI . DEF
DEFHI . DEF
DEFIJ . DFG
DEJ . DFG
DG
GL
GHKL
.............................................................................
76861'
/TIME
If more proof was needed, all one had to do was wait until the open-
ing of July 6. An initiating selling extreme developed on the opening
under the previous day's value area. This confirmed the presence of sell-
ers and difficulty ahead. Thus, the Market Profile@tools confirmed the
bearish implications of the shooting star.
There are some interesting similarities between Market con-
cepts and candlesticks. Wider value areas in Market Profile@usually rep-
resent facilitation of trade and, as such, increase the probability for price
trend continuation. Thus, in an one would like to see widening
value areas. Likewise, with the candlesticks, one would like to see a rally
via a series of longer and longer white real bodies in order to confirm the
power behind the move.
Shrinking value areas in Market Profile@reflect less facilitation of
trade and thus less certainty of a continuation of the price move. So it is
with the advance block or stalled patterns. In those forma-
tions, the trend is still up but it takes place by means of shrinking white
real bodies. These formations indicate that the prior momentum is run-
ning out of steam.
What about a star in candlesticks? A short real body in an or
downtrend would be a sign of decreasing vigor by the bulls (a star in an
or the bears (a star in a downtrend). So would a small value
area after a strong advance (or decline). The small value area would
reflect a lack of trade facilitation. They could be a harbinger of a trend
change. A hammer's lower shadow might be formed due to a buying
extreme in which lower prices induce an influx of buyers. A shooting
star's long upper shadow could be the result of a selling extreme in
which higher prices attracted strong selling.
CANDLESTICKS
WITH OPTIONS
there is a lid that does not fit, there is a lid that does"
OPTIONS BASICS
Before discussing how candlesticks can be used with options, we'll want
to spend some time on option basics. The five factors needed to figure
the theoretical value of a futures options price are the exercise price, the
time to expiration, the price of the underlying instrument, volatility, and, to a
minor extent, short-term interest rates. Three of these variables are known
(time until expiration, exercise price, and short-term interest rates). The
two components that are not known (the forecasted price of the under-
lying instrument and its volatility) have to be estimated in order to fore-
cast an option price. One should not underrate the importance that
volatility plays in option pricing. In fact, at times, a change in volatility
can have a stronger impact on option premiums than a change in price
of the underlying contract.
Consider as an illustration an at-the-money $390 gold call with 60
days until expiration. If this option has a volatility level of its the-
oretical price would be $1,300. At a 15% volatility, this same option
would be theoretically priced at $1,000. Thus, volatility must always be
considered since it can so forcefully affect the option premium.
Volatility levels provide the expected range of prices over the next
year (volatilities are on an annual basis). Without getting into the math-
ematics, a volatility of 20% on, say, gold suggests that there is a 68%
probability that a year from now gold's price will remain within plus or
minus 20% of its current price. And there would be a 95% chance that a
year hence golds's price will remain within plus or minus two times the
volatility (that is, two times 20% or 40%) of its price now. For example,
if gold is at $400 and volatility is at there is a 68% chance that its
price after one year will be between $320 and $480 (plus or minus 20% of
$400) and a 95% probability that it will be between $240 and $560 (plus
or minus 40% of $400). Keep in mind that these levels are based on
probabilities and that at times these levels are exceeded.
The greater the volatility the more expensive the option. This is due
to at least three factors. First, from the speculators' point of view, the
greater the volatility, the greater the chance for prices to move into the
money (or further into the money). Second, from a hedger's perspective,
higher price volatility equals more price risk. Thus, there is more reason
to buy options as a hedging vehicle. And, third, option sellers also
require more compensation for higher perceived risk. All these factors
will buoy option premiums.
There are two kinds of volatility: historic and implied. Historic volatil-
ity is based on past volatility levels of the underlying contract. It is usu-
ally calculated by using daily price changes over a specified number of
business days on an annualized basis. In the futures markets, 30
Candlesticks Options 269
days are the most commonly used calculations. Just because a futures
contract has a 20-day historic volatility of 15% does not mean it will
remain at that level during the life of the option. Thus, to trade options
it is necessary to forecast volatility. One way to do this is by having the
market do it for you. And that is what implied volatility does. It is the
market's estimate of what volatility of the underlying futures contract
will be over the options's life. This differs from historic volatility in that
historic volatility is derived from prior price changes of the underlying
contract.
Implied volatility is the volatility level that the market is implying
(hence its name). Deriving this number involves the use of a computer
but the theory behind it is straightforward. To obtain the option's
implied volatility, the five inputs needed are the current price of the
futures contract, the option's strike price, the short-term interest rate,
the option expiration date, and the current option price. If we put these
variables into the computer, using an options pricing formula, the com-
puter will spit back the implied volatility.
Thus, we have two volatilities-historic, which is based on actual
price changes in the futures, and implied, which is the market's
guess of what volatility will be from now until the option expires. Some
option traders focus on historic volatility, others on implied volatility,
and still others compare historic to implied.
17.50
15.88
12.58
Thus, for those who were riding volatility on the way up, the harami
cross could have been viewed as a signal to expect an end to the prior
steep trend and, consequently, the possibility of a decline in volatility.
Based on experience, and for various other reasons, candlestick sig-
nals seem to work better with historic volatility than with implied vola-
tility. Nonetheless, as shown in Exhibit 18.2, candlesticks, at times, can
be useful instruments to assist in forecasting short-term moves in
implied volatility. A sharp rally developed in January. During this
phase, implied volatility ascended along an For much of Febru-
ary, sugar was bounded in a to range. In this period of rela-
tively quiet trading, volatility shrank.
On February 26, a hammer appeared (area A in Exhibit In
addition, this hammer session's lower shadow broke under the support
area from late January. This new low did not hold. This demonstrated
that the bears tried to take control of the market, but failed. Three trad-
ing day's later, on March 2, (area B in Exhibit the white candle-
lower shadow successfully maintained the lows of mid-January.
In addition, the March 2 low joined with the hammer to complete a
tweezers bottom. The combination of all these bottom reversal indicators
sent a powerful signal that a solid base had been built. A significant rally
was thus possible. As shown in Exhibit volatility at areas A and
B were at relatively low levels. With a possible strong price rally (based
on the confluence of bottom reversal signals discussed above) and low
volatility, one could expect any price rallies to be mimicked with expand-
ing volatility. That is what unfolded.
Another use for options and candlesticks is for the risky proposition
of bottom and top picking. Just about any book on trading strategies
warns against this. But let's face it, we all occasionally attempt it. This is
an example' of how the limited risk feature of options may allow one to
place a trade too risky for an outright future. Exhibit 18.3 is an example
of a trade I recommended. I would not have made such a recommenda-
tion without the limited risk feature of options.
Cocoa was in a major bull market that started in November 1989 at
$900. This exhibit shows the last three waves of an Elliott Wave count.
The top of wave 3 was accompanied by the shooting star and the bottom
of wave 4 by the bullish piercing line. Based on a Fibonacci ratio, there
is a wave 5 target near $1,520. Thus, near $1,520 one should start look-
ing for candlestick confirmation of a top. And in late May, a bearish
engulfing pattern emerged after the market touched a high of $1,541.
This was close to the Elliott Wave count of $1,520 which signaled a pos-
sible top.
I could not resist such a powerful combination of an Elliott fifth wave
and a bearish engulfing pattern! Options to the rescue! I recommended
272 The Rule of Multiple Technical Techniques
,. . .
I
Mar
Source:
EXHIBIT 18.2. Sugar-July 1990 (a) Daily (Candlesticks with Options) and (b) Implied
Volatility
Candlesticks with Options 273
a buy of the $1,400 puts (because the major trend was still up, I would
have been more comfortable liquidating longs, but unfortunately I was
not long during this rally). If I was wrong about a top and cocoa gapped
higher (as it did in mid-May), increased volatility could help mitigate
adverse price action. As it turned out, this bearish engulfing pattern and
Elliott fifth wave became an important peak.
HEDGING WITH
CANDLESTICKS
futures or calls. If prices move up, the higher price needed to purchase
the cash commodity would be offset, at least partially, by the gains in the
futures or options positions.
In many instances, the underlying cash position may not be fully
hedged at one price. The hedger may intend to scale into the hedge
position. Thus, the question a hedger must frequently address is what
percent of his exposure should be hedged. Candlestick techniques can
help answer this question. By exploring the following examples, it
should offer insight into how candlesticks can be used to furnish clues
as to when to adjust the hedged portion of the cash position.
Candlestick techniques can also be used for those who are 100%
hedged. For instance, say that you are a corn farmer and the corn
market is trending against your cash position (corn prices are falling).
This should mean, as a short hedger, your futures hedge position is
profitable. If a strongly bullish candlestick reversal formation appears,
and you believe prices will rally, you might want to ease back on the
hedge. In some instances, early lifting of a profitable hedge may improve
cash flow. However, no hedge should be viewed as a strategy to gener-
ate profits.
As shown in Exhibit 19.1, an explosive rally unfolded in 1988 as
prices nearly doubled from $6 to $11. In mid-1988, near the peak of the
rally, the candlesticks gave a reversal signal a the bearish engulfing
pattern. This would have not been seen as a Western top reversal forma-
tion using traditional Western techniques. A top reversal formation
would have required a new high for the move and then a close under the
prior week's close. The black candlestick session of the engulfing pattern
did not create a new high. Thus, while not a Western top reversal pat-
tern, it was a reversal formation with candlesticks. This bearish engulf-
ing pattern, for a soybean farmer, could have been used as a warning to
either initiate short hedge positions or, if not already fully hedged, to
raise the percentage hedged.
In 1989, another bearish engulfing pattern warned of a top (this also
was not seen as a top reversal pattern using Western technicals). Short
hedgers could have placed hedge positions here. The window a few
weeks later was another bearish candlestick signal which would have
told short hedgers to add to their short hedge position (if not already
100% hedged). October's bullish piercing pattern could have been used
as a signal to ease back on short-hedge positions. For those looking at a
long hedge, the piercing pattern could be used as a signal to initiate such
a hedge.
As illustrated in Exhibit 19.2, a confluence of bullish candlestick sig-
nals emerged the last week in September and the first week in October
in 1988. The most obvious was the bullish engulfing pattern. These two
weeks also created a tweezers bottom. Additionally, the white candle-
stick was a bullish belt hold which closed at its high. It also engulfed the
prior five candlesticks. For long hedgers, these would have been signs to
start, or add to, their hedges.
In early 1989, the harami cross (an important reversal signal), could
have been used by long hedgers to scale back on hedges. Also, the fact
that this harami cross was formed by a long white candlestick and a doji
would indicate that the market was in trouble. As discussed in Chapter
8, a doji after a long white real body is often a sign of a top.
Exhibit 19.3 explores how a long crude oil hedger could use a candle-
stick signal to get an early signal of a bottom reversal and then use West-
ern technical tools to confirm. a bottom and add to the long hedge. The
first tentative bottom reversal clue came with a hammer on July 5 . Since
this hammer session gapped under the prior lows, it should have not
been taken as a bullish sign unless it was confirmed by other bullish
indicators during the next few sessions. The next clue was the bullish
engulfing pattern which provided bullish confirmation to the hammer.
This engulfing pattern was also an extra important bottom reversal sig-
nal. This is because the white candlestick on July 9 engulfed not one but
two black real bodies. At this point, crude oil end users (that is, long
278 The Rule of Multiple Technical Techniques
Bullish
Engulfing
I
Jul Oct Jan Jul Oct
EXHIBIT 19.2. Cocoa-Weekly
(Candlesticks with Hedging) Source: Commodity Trend Service"
,
. . ..
..
.:. 1900
..
1800
.. Bullish .
7/24/90 Engulfing
Pattern
-200
. . . .
" A prudent man has more than one string to his bow"
On June 13, I recommended a short at $6.27. The stop was above the
neckline at $6.35. The target was $5.90 (this was based on a support area
earlier in the year). The short position looked good for the rest of the
week as the market tumbled. Then on June 18, the market shouted it
was time to cover shorts. The clues were:
The tall white candlestick on June 18 engulfed the prior black real
body. These two candlesticks formed a bullish engulfing line.
The opening on June 18 made a new low for the move by punching
under the early June lows. However, prices bounced right back above
this early June low. This created a spring in which the lows were bro-
ken and not held. Based on this spring, I had targeted a retest of the
upper end of the trading range at $6.30.
The candlestick on the June 18 was a strong bullish belt-hold line.
The new price lows were not validated by lower stochastic levels. This
was a positive divergence and it demonstrated that the bears were
losing control.
How I Have Used Candlesticks 283
. A = -200
4 .. ..
..
Resistance'
Zone 9000
.. ..
.. ..
Hararni
6/26/90 . .. ..
. 8825
8825 .. ..
. L = 8615
8650 May
. Hammer
1677 .
.-0= 1678 .
1663 .
1672 .
6/18 6/25 2
This level became resistance. Once the market backed down from near
on the brief rally, I knew the bears were still in charge. The move
on June 20 opened a window. This implied another leg lower. On June
21, when the market failed to move above the window's resistance area,
I recommended a short at with a stop above the June 20 high of
The objective was for (this was based on support earlier
in the year). The market then backed off.
At the time of the chart shown in Exhibit 20.3, crude oil was in a bear
market. So I was looking to short on rallies. A rally started on June 21
with the hammer (note how I did not use this bullish hammer as a time
to buy. Why? Because the major trend was down). Then, a few days
after the hammer, back-to-back bearish engulfing patterns emerged. At
that time, I suggested a short sale. There was a small window opened
between June 15 and June 18. In the belief that this window would be
resistance, I put a stop slightly above it at $17.65. The target was a retest
near the hammer lows. A late rally on June 25 quickly backed off from the
resistance area made by the bearish engulfing pattern and the window.
Exhibit 20.4 shows that on June 4 a hammer and a successful hold-
How I Have Used Candlesticks 285
2700
and Spring
2600
.. . . .
, 2810
2776
' 2806 'May 'Jun
ing, on a close, of the late April, mid-May lows of $2.65. It was also a
bullish spring since prices made new lows for the move but these new
lows failed to hold.
Although this was a bullish hammer I needed confirmation the next
day before I could recommend to buy since the market was holding the
$2.65 support so tenuously. The day after this hammer the market
proved itself by opening higher. I recommended a buy at $2.68 with a
stop under the support line shown at $2.64. The target was for a test of
the downward sloping resistance line at $2.79. The strong session after
the hammer day made this into a morning star pattern.
Although the trade in Exhibit 20.5 was not profitable, it is a good
example of how candlesticks can help to provide good trade location.
The price action on Friday, March 2 gave some buy signals. First, it was
a hammer line. Next, this hammer's small real body was inside the prior
tall real body creating a harami. This harami meant that the prior minor
downtrend was over. Finally, the hammer session's lows took out sup-
port from early February, yet these new lows could not hold. In other
words, the bears tried to break the market and they could not. It was
time to step in and buy.
286 The Rule of Multiple Technical Techniques
CANDLESTICK TERMS
AND VISUAL DICTIONARY
Advance block-a variation on three white soldiers in which the last two
soldiers white real bodies) display weakening upside drive. This
weakness could be in the form of tall upper shadows or progressively
smaller real bodies. It signifies a diminution of buying force or an
increase in selling pressure.
Advance Block
290 Glossary A: Candlestick Terms and Visual Dictionary
Belt-hold line-there are bullish and bearish belt holds. A bullish belt hold
is a tall white candlestick that opens on its low. It is also called a white
opening shaven bottom. At a low price area, .this is a bullish signal. A
bearish belt hold is a long black candlestick which opens on its high.
Also referred to as a black opening shaven head. At a high price level, it
is considered bearish.
Bullish Bearish
Candlestick
Line
Shadow
Counterattack Lines
Glossary A: Candlestick Terms and Visual Dictionary 291
Dark-Cloud Cover
Doji
Doji star-a doji line which gaps from a long white or black candlestick.
An important reversal pattern with confirmation during the next ses-
sion.
Doji Star
Evening Star
Gapping Play
High Low
Price Price
Window
real body can be white or black. Most often the second real body is
the opposite color of the first real body.
cross-a harami with a doji on the second session instead of a
small real body. An important top (bottom) reversal signal especially
after a tall white (black) candlestick line. It is also called a petrifying
pattern.
Harami Cross
I I
High Waves
High wave-a candlestick with a very long upper or lower shadow and a
short real body. A group of these can foretell a market turn.
High-price gapping play-see "Gapping plays."
In-Neck In-neck line-a small white candlestick in a downtrend whose close is a
slightly above the previous black candlestick's low of the session.
After this white candlestick's low is broken, the downtrend should
continue. Compare to on-neck line, thrusting line, and piercing pat-
tern.
Inverted hammer-following a downtrend, this is a candlestick line that
has a long upper shadow and a small real body at the lower end of
the session. There should be no, or very little, lower shadow. It has
the same shape as the bearish shooting star, but when this line occurs
in a downtrend, it is a bullish bottom reversal signal with confirma-
tion the next session a white candlestick with a higher close or
a higher opening).
Inverted Hammer
Separating Lines
Bullish Bearish
Shadows-the thin lines above and below the real body of the candlestick
line. They represent the extremes of the day. The lower shadow is
the line on the bottom of the real body. The bottom of the lower
shadow is the low of the session. The upper shadow is the line on
top of the real body. The top of the upper shadow is the high of the
session. See the illustration under "Candlestick lines and charts."
Shaven bottom-a candlestick with no lower shadow.
Shaven head-a candlestick with no upper shadow.
Shooting Star Shooting star-a candlestick with a long upper shadow with little, or no
lower shadow, and a small real near the lows of the session that
arises after an It is a bearish candlestick signal in an
gapping lower are bearish since they are viewed as temporary short
covering. Gapping side-by-side lines are very rare.
Side by Side White Lines
Three Buddha patterns-A three Buddha top is the same as the Western
head and shoulders top. In Japanese terms, the three Buddha top is
a three mountain top in which the central mountain is the tallest. An
inverted three Buddha is the same as the Western inverted head and
shoulders. In Japanese terminology, it is a three river bottom in
which the middle river is the longest.
Three Buddha
Three river bottom-when the market hits a bottom area three times.
When the peak of the intervening valleys is exceeded by a white can-
dlestick or with a gap it is confirmation that a bottom has been put in
place.
Three River Bottom
Thrusting line-a white candlestick which closes in the prior black real
body, but still under the middle of the prior session's real body. The
thrusting line is stronger than an in-neck line, but not as strong as a
piercing line. In a downtrend, the thrusting line is viewed as bearish
(unless two of these patterns appear within a few days of each other).
As part of a rising market it is considered bullish.
Thrusting Line
Towers-there is a tower top and tower bottom. The tower top, a top
reversal formation, is comprised of a tall white candlestick followed
by congestion and then one or more long black candlesticks. It is a
pattern which looks like it has towers on both sides of the congestion
Towers
Bottom
300 Glossary A: Candlestick and Visual Dictionary
Tweezers
Tops Bottoms
I I
Unique three river bottom-a rare type of bottom comprised of three lines.
River Bottom
The first is a long black real body, the second is a hammer like ses-
sion with a black real body which makes a new low, and the third
candlestick is a small real body.
I Upper shadow-see "Shadows."
Upside gap tasuki-see "Tasuki gaps.
Upside gap two crows-a three candlestick pattern. The first line is a long
T w o Crows
white candlestick which is followed by a black real body that gaps
higher. The third session is another black real body which opens
above the second session's open and closes under the second ses-
sion's close. It is a top reversal signal.
Window-the same as a Western gap. Windows are continuation pat-
terns. When the market rallies and opens a window, there should be
a pullback to that window. The window should be support. If a
Windows
I,
In an In a Downtrend
Glossary A: Candlestick Terms and Visual Dictionary 301
AMERICAN
TECHNICAL TERMS
This glossary clarifies the Western technical terms used in this book.
It is not meant to be comprehensive or detailed because this book's focus
is on Japanese candlesticks and not Western technical tools.
Bar chart-a graphic representation of price activity. The high and low of
the session define the top and bottom of a vertical line. The close for
the period is marked with a short horizontal bar attached to the right
of the vertical line. The open is marked with a short horizontal bar
attached to the left of the vertical line. Price is in on the vertical scale;
time is on the horizontal scale.
Blow-offs-a top or bottom reversal. Blow-offs occur after an extended
move. Prices, usually with very high volume, sharply and quickly
thrust strongly in the direction of the preceding trend. If the market
reverses after this action, it is a blow-off.
Breakaway gap-when prices gap away from a significant technical area
a trendline or a congestion zone).
Breakout-overcoming a resistance or support level.
Change of polarity-when old support converts to new resistance or when
old resistance converts to new support.
Confirmation-when more than one indicator substantiates the action of
another.
Congestion zone or band-a period of lateral price action within a relatively
narrow price band.
Consolidation-the same as a congestion zone. Consolidation, however,
has the implication that the prior trend should resume.
304 Glossary B: American Technical
Paper trading-not trading with real money. All transactions are only
imaginary with a record of profit and loss on paper.
Pennant-see "Flag."
Positive crossover-see "Crossover.
Positive divergence-see "Divergence.
Protective stop-a means of limiting losses if the market moves against
your position. If your stop level is reached, your position is automat-
ically offset at the prevailing price.
Rally-an upward movement of prices.
Reaction-a price movement opposite to the prevailing trend.
Relative Strength Index-an oscillator developed by Welles Wilder. The
RSI compares the ratio of up closes to down closes over a specified
time period.
Resistance level-a level where sellers are expected to enter.
price reaction from the prior move in percentage terms.
The more common retracement levels are and 62%.
Reversal session-a session when a new high (or low) is made for the
move and the market then closes under (or above) the prior session's
close.
Reversal indicator-see "Trend reversals."
Selling climax-when price push sharply and suddenly lower on heavy
volume after an extended decline. If the market reverses from this
sharp it is viewed as a selling climax.
Selloff-a downward movement of prices.
Simple moving average-a method of smoothing price data in which prices
are added together and then averaged. It is a "moving" average
because the average moves. As new price data is added the oldest
data is dropped.
Spring-when prices break under the support of a horizontal congestion
band and then springs back above the "broken support" area. This is
bullish.
Stochastics-an oscillator that measures the relative position of the closing
price as compared to its range over a chosen period. It is usually
comprised of the faster moving line and the slower moving
line.
Support level-an area where buyers are expected to enter.
Tick VolumeTM-the number of trades per given intra-day time period.
Glossary B: American Technical Terms 307
Time filter-Prices have to stay above, or below, a certain price area for a
specific time to confirm that an important technical area has been
broken. For example, the market might have to close above a broken
resistance level for two days before a long position is placed.
Trading range-when prices are locked between horizontal support and
horizontal resistance levels.
Trend-the market's prevalent price direction.
Trend reversals-also called reversal indicators. This is a misleading term.
More appropriate, and more accurate, would be the term "trend
change indicator." It means the prior trend should change. It does
not mean prices are going to reverse.
Trendline-a line on a chart that connects a series of higher highs or
lower lows. At least two points are needed to draw a trendline. The
more often it is tested, and the greater the volume on the tests, the
more important the trendline.
Upgap-a gap which pushes prices higher.
Upthrust-when prices break above a resistance line from a laterally trad-
ing zone. If these new highs fail to hold and prices pull back under
the "broken" resistance line it is an upthrust. It is a bearish signal.
Uptrend-a market that is trending higher.
V bottom or top-when prices suddenly reverse direction forming a price
pattern that looks like the letter V for a bottom or an inverted V for
a top.
Volume-the total of all contracts traded for a given period.
Weighted moving average-a moving average in which each of the previous
prices is assigned a weighting factor. Usually, the most recent data is
the more heavily weighted.
BIBLIOGRAPHY
Analysis of Stock Price in Japan, Tokyo, Japan: Nippon Technical Analysts Association 1986.
Buchanen, Daniel Crump, Japanese Proverbs and Sayings, Oklahoma City: University of Oklahoma
Press, 1965.
Colby, Robert W. and Thomas A. Meyers, The Encyclopedia of Technical Market Indicators,
wood, Dow Jones-Irwin, 1988.
Dilts, Marion May, The Pageant of Japanese History, New York: David 1963.
Drinka, Thomas and Robert L. "Market Profile and Market Logic," Technical Analysis
of Stocks and Commodities, December 1987, pp.
Edwards, Robert D. and John Technical Analysis of Stock Trends, 5th ed., Boston: John
1966.
Hill, Julie Skur, "That's Not What I Said," Business Tokyo, August 1990, pp.
Hirschmeier, Johannes and Tsunehiko Yui, Development of Japanese Business 1600-1973, Cam-
bridge, MA: Harvard University Press 1975.
Hoshii, Kazutaka, Kabuka Chato wo no Hon ( A Book for Those Reading Stock Charts
for the First Time), Tokyo, Japan: Asukashuppansha, 1990.
Ikutaro, Gappo, Kabushikisouba no Technical Bunseki (Stock Market Technical Analysis), Tokyo, Japan:
Nihon Keizai Shinbunsha, 1985.
Ishii, Katsutoshi, Kabuka Chato no Tashikana Yomikata Sure Way to Read Stock Charts), Tokyo,
Japan: Jiyukokuminsha, 1990.
Kaufman, Perry J. The New Commodity Trading Systems and Methods, New York: John Wiley and
Sons, 1987.
Keisen Kyoshitsu Part 1 (Chart Classroom Part Tokyo, Japan: Toshi Rader, 1989.
Kroll, Stanley, Kroll on Futures Trading, Homewood, Dow Jones-Irwin, 1988.
Masuda, Koh, ed. New School Japanese-English Dictionary, Tokyo, Japan: Kenkyusha,
1968.
Murphy, John J., Technical Analysis of the Futures Markets, New York: New York Institute of
Finance, 1986.
Nihon Keisenshi (The History of Japanese Charts), Chapter 2 by Kenji Oyama, pp. 90-102, Tokyo,
Japan: Nihon Keisai Shimbunsha, 1979.
Okasan Keisai Kenkyusho, Jissen Kabushiki (Introduction to Stock Charts), Tokyo,
Japan: Diamond-sha, 1987.
Sakata Goho Five Rules are Wind, Forest, Fire and Mountain), Tokyo, Japan:
Nihon Shoken Shimbunsha, 1969.
Schabacker, Richard W., Technical Analysis and Stock Market Profits, New York: The Schabacker
Institute.
310 Bibliography
Seidensticker, Edward G., Even Monkeys Fall Trees and Other Japanese Proverbs, VA:
Charles E. Tuttle, 1987.
Seward, Jack, Japanese in Action, New York: Weatherhill, 1983.
Shimizu, Seiki, The Japanese Chart of Charts, trans. Gregory S. Nicholson, Tokyo, Japan: Tokyo
Futures Trading Publishing Co., 1986.
Sklarew, Arthur, Techniques of a Professional Commodity Chart Analyst, New York: Commodity
Research Bureau, 1980.
Smith, Adam, The Money Game, New York: Random House, 1968.
Tamarkin, Robert, The New Gatsbys, Chicago, Bob Tamarkin, 1985.
Taucher, Frank, Commodity Trader's Almanac, Tulsa: Market Movements, 1988.
Technical Traders Bulletin, January 1990, May 1990, June 1990, Rolling Hill Estates, CA: Island View
Financial Group Inc., 1990.
Wilder, J. Welles, New Concepts in Technical Trading Systems, Greensboro, NC: Trend Research,
1978.
Yoshimi, Toshihiko, Yoshimi Toshihiko no Chato Kyoshitsu (Toshihiko Chart Classroom),
Tokyo, Japan: Nihon Chart, 1989.
Index
Trendlines, and candlesticks, Upside-gap two crows, 98-101, 300 relative strength index, 231
207 Upthrusts, 193-200 tweezers top pattern, 93
change of polarity principle, Upward sloping resistance line, upthrusts and scouting parties
201-7 192 pattern, 200
springs and upthrusts, 193-200 White candlestick, 24
support and resistance lines, v White opening shaven bottom, 94
185-93 Value area, 260 Windows, 110, 119-34, 300-301
Tri-star, 162-63, 300 Visual dictionary, 289-301 patterns which include win-
Tweezers tops and bottoms, 88-94, Volatility, 268-69 dows, 129-34
300 Volume, and candlesticks, 241-48 Wyckoff, Richard, 193
on balance volume, 244-45
u tick volume, 245-48 Y-z
Unique three river bottom pattern, V reversals, 118 Yen
116-17, 300 bearish counterattack line pat-
Unleaded gas W tern of, 105
bearish counterattack line pat- Wave principle, 253 piercing pattern of, 53
tern, 105 Weighted moving average, 216 shooting star pattern of, 71
spring with candlestick pattern, Wheat, 166-67 Yin and yang lines, 24, 301
198 confluence of candlesticks, 181 Yodoya Keian, 14-15
side-by-side white lines, doji at tops pattern, 152 Yorikiri, 95
134 harami pattern, 84 Yo-sen, 24
Upper shadow, 23 morning doji star pattern, 66 Zinc, shooting stars and inverted
Upside-gap tasuki, 129-30 piercing pattern, 51-52 hammer patterns of, 77